Skip to content

Commit

Permalink
fix: fix clip tokenizer (#383)
Browse files Browse the repository at this point in the history
  • Loading branch information
stduhpf authored Sep 2, 2024
1 parent e410aeb commit 14206fd
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion clip.hpp
Original file line number Diff line number Diff line change
Expand Up @@ -388,7 +388,7 @@ class CLIPTokenizer {
std::string token_str = token.str();
std::u32string utf32_token;
for (int i = 0; i < token_str.length(); i++) {
char b = token_str[i];
unsigned char b = token_str[i];
utf32_token += byte_encoder[b];
}
auto bpe_strs = bpe(utf32_token);
Expand Down

0 comments on commit 14206fd

Please sign in to comment.