CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-30 08:42:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama.cpp : split llama_context_params into model and context params (#3301)

Browse Source 
* llama.cpp : split llama_context_params into model and context params

ggml-ci

* fix metal build

* fix freq_base/scale default to model value

* llama-bench : keep the same model between tests when possible

* move n_threads to llama_context_params, add n_threads_batch

* fix mpi build

* remove kv_size(), cuda scratch fixes

* remove low-vram option

* add n_threads_batch to system info, refactor to get_system_info()

* add documentation about --threads-batch to the READMEs

* llama-bench fix

* main : fix rope freq/scale warning

* llama.cpp : add llama_get_model
common : add llama_tokenize from model

* remove duplicated ctx/model functions

ggml-ci

* cuda : print total VRAM used
This commit is contained in:
slaren
2023-09-28 21:42:38 +02:00
committed by GitHub
parent 0512d66670
commit 16bc66d947
27 changed files with 713 additions and 633 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
examples/embedding/embedding.cpp
View File

@@ -42,17 +42,18 @@ int main(int argc, char ** argv) {
return 1;
}
const int n_ctx_train = llama_n_ctx_train(ctx);
if (params.n_ctx > n_ctx_train) {
const int n_ctx_train = llama_n_ctx_train(model);
const int n_ctx = llama_n_ctx(ctx);
if (n_ctx > n_ctx_train) {
fprintf(stderr, "%s: warning: model was trained on only %d context tokens (%d specified)\n",
__func__, n_ctx_train, params.n_ctx);
__func__, n_ctx_train, n_ctx);
}
// print system information
{
fprintf(stderr, "\n");
fprintf(stderr, "system_info: n_threads = %d / %d | %s\n",
params.n_threads, std::thread::hardware_concurrency(), llama_print_system_info());
fprintf(stderr, "%s\n", get_system_info(params).c_str());
}
int n_past = 0;
@@ -70,15 +71,15 @@ int main(int argc, char ** argv) {
fprintf(stderr, "\n");
}
if (embd_inp.size() > (size_t)params.n_ctx) {
if (embd_inp.size() > (size_t)n_ctx) {
fprintf(stderr, "%s: error: prompt is longer than the context window (%zu tokens, n_ctx = %d)\n",
__func__, embd_inp.size(), params.n_ctx);
__func__, embd_inp.size(), n_ctx);
return 1;
}
while (!embd_inp.empty()) {
int n_tokens = std::min(params.n_batch, (int) embd_inp.size());
if (llama_decode(ctx, llama_batch_get_one(embd_inp.data(), n_tokens, n_past, 0), params.n_threads)) {
if (llama_decode(ctx, llama_batch_get_one(embd_inp.data(), n_tokens, n_past, 0))) {
fprintf(stderr, "%s : failed to eval\n", __func__);
return 1;
}
@@ -86,8 +87,8 @@ int main(int argc, char ** argv) {
embd_inp.erase(embd_inp.begin(), embd_inp.begin() + n_tokens);
}
const int n_embd = llama_n_embd(ctx);
const auto embeddings = llama_get_embeddings(ctx);
const int n_embd = llama_n_embd(model);
const auto * embeddings = llama_get_embeddings(ctx);
for (int i = 0; i < n_embd; i++) {
printf("%f ", embeddings[i]);