root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-28 12:24:35 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Update gptneox-main.cpp

Browse Source
This commit is contained in:
klosax 2023-08-07 22:28:43 +02:00 committed by GitHub
parent 8083ae347a
commit 65559a23c8
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 0 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
gptneox-main.cpp
View File

@ -443,9 +443,6 @@ bool gpt_neox_model_load(const std::string & fname, gpt_neox_model & model, gpt2
// load vocab
{
// TODO: implement a better bpe tokenizer, utilizing merges and handles unicode
auto & hparams = model.hparams;
int keyidx = gguf_find_key(ggufctx, "tokenizer.ggml.model");
@ -484,11 +481,6 @@ bool gpt_neox_model_load(const std::string & fname, gpt_neox_model & model, gpt2
for (size_t i = 0; i < hparams.n_vocab; i++) {
std::string word = gguf_get_arr_str(ggufctx, tokens_keyidx, i);
// TEMP until a better bpe tokenizer is implemented
// word = replace(word, "Ġ", " ");
// word = replace(word, "Ċ", "\n");
// printf("token %d = '%s'\n",i,word.c_str() );
vocab.token_to_id[word] = i;
@ -1054,7 +1046,6 @@ int main(int argc, char ** argv) {
{
const int64_t t_start_sample_us = ggml_time_us();
// id = sample_top_k_top_p(vocab, logits.data() + (logits.size() - n_vocab), top_k, top_p, temp, repeat_last_n, repeat_penalty, rng);
id = sample_top_k_top_p_repeat(vocab, logits.data() + (logits.size() - n_vocab), last_n_tokens.data(), last_n_tokens.size(), top_k, top_p, temp, repeat_last_n, repeat_penalty, rng);
last_n_tokens.erase(last_n_tokens.begin());