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

llama : disambiguate API

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-09-02 10:06:42 +03:00
parent 8f1d81a0b6
commit 086e7f6ebc
No known key found for this signature in database
GPG key ID: BF970631944C16B7
9 changed files with 22 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

4
examples/embedding/embedding.cpp
View file

@ -31,7 +31,7 @@ static void batch_add_seq(llama_batch & batch, const std::vector<int32_t> & toke
}
static void batch_decode(llama_context * ctx, llama_batch & batch, float * output, int n_seq, int n_embd, int embd_norm) {
const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
const enum llama_pooling_type pooling_type = llama_get_pooling_type(ctx);
const struct llama_model * model = llama_get_model(ctx);
// clear previous kv_cache values (irrelevant for embeddings)
@ -114,7 +114,7 @@ int main(int argc, char ** argv) {
const int n_ctx_train = llama_n_ctx_train(model);
const int n_ctx = llama_n_ctx(ctx);
const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
const enum llama_pooling_type pooling_type = llama_get_pooling_type(ctx);
if (llama_model_has_encoder(model) && llama_model_has_decoder(model)) {
fprintf(stderr, "%s: error: computing embeddings in encoder-decoder models is not supported\n", __func__);

2
examples/perplexity/perplexity.cpp
View file

@ -796,7 +796,7 @@ static void hellaswag_score(llama_context * ctx, const gpt_params & params) {
size_t hs_task_count = prompt_lines.size()/6;
fprintf(stderr, "%s : loaded %zu tasks from prompt.\n", __func__, hs_task_count);
const bool is_spm = llama_vocab_type(llama_get_model(ctx)) == LLAMA_VOCAB_TYPE_SPM;
const bool is_spm = llama_get_vocab_type(llama_get_model(ctx)) == LLAMA_VOCAB_TYPE_SPM;
fprintf(stderr, "================================= is_spm = %d\n", is_spm);
// The tasks should be randomized so the score stabilizes quickly.

2
examples/retrieval/retrieval.cpp
View file

@ -162,7 +162,7 @@ int main(int argc, char ** argv) {
const int n_ctx_train = llama_n_ctx_train(model);
const int n_ctx = llama_n_ctx(ctx);
const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
const enum llama_pooling_type pooling_type = llama_get_pooling_type(ctx);
if (pooling_type == LLAMA_POOLING_TYPE_NONE) {
fprintf(stderr, "%s: error: pooling type NONE not supported\n", __func__);
return 1;

2
examples/server/server.cpp
View file

@ -2450,7 +2450,7 @@ struct server_context {
json model_meta() const {
return json {
{"vocab_type", llama_vocab_type (model)},
{"vocab_type", llama_get_vocab_type(model)},
{"n_vocab", llama_n_vocab (model)},
{"n_ctx_train", llama_n_ctx_train (model)},
{"n_embd", llama_n_embd (model)},

4
examples/speculative/speculative.cpp
View file

@ -82,10 +82,10 @@ int main(int argc, char ** argv) {
model_dft = llama_init_dft.model;
ctx_dft = llama_init_dft.context;
const bool vocab_type_tgt = llama_vocab_type(model_tgt);
const bool vocab_type_tgt = llama_get_vocab_type(model_tgt);
LOG("vocab_type tgt: %d\n", vocab_type_tgt);
const bool vocab_type_dft = llama_vocab_type(model_dft);
const bool vocab_type_dft = llama_get_vocab_type(model_dft);
LOG("vocab_type dft: %d\n", vocab_type_dft);
if (vocab_type_tgt != vocab_type_dft) {