CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-06 09:46:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

only rm when params.escape, rm space if possible which is added back or rm added space token

Browse Source
This commit is contained in:
vvhg1
2023-10-07 09:08:30 +02:00
parent b4046aabbf
commit 0526560759
2 changed files with 13 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
examples/server/server.cpp
View File

@@ -344,11 +344,16 @@ struct llama_server_context
void loadInfill()
{
params.input_suffix.erase(0, params.input_suffix.find_first_not_of(" "));
bool suff_rm_leading_spc = params.escape;
if (suff_rm_leading_spc && params.input_suffix.find_first_of(" ") == 0 && params.input_suffix.size() > 1) {
params.input_suffix.erase(0, 1);
suff_rm_leading_spc = false;
}
auto prefix_tokens = tokenize(params.input_prefix, false);
auto suffix_tokens = tokenize(params.input_suffix, false);
const int space_token = 29871;
if (params.escape && suffix_tokens.size() > 1 && suffix_tokens[0] == space_token) {
if (suff_rm_leading_spc && suffix_tokens[0] == space_token) {
suffix_tokens.erase(suffix_tokens.begin());
}
prefix_tokens.insert(prefix_tokens.begin(), llama_token_prefix(ctx));