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 : fix time complexity of string replacement (#9163)

Browse Source 
This change fixes a bug where replacing text in a very long string could
cause llama.cpp to hang indefinitely. This is because the algorithm used
was quadratic, due to memmove() when s.replace() is called in a loop. It
seems most search results and LLM responses actually provide the O(n**2)
algorithm, which is a great tragedy. Using a builder string fixes things
This commit is contained in:
Justine Tunney
2024-08-25 23:09:53 -07:00
committed by GitHub
parent 93bc3839f9
commit 436787f170
3 changed files with 30 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
common/common.cpp
View File

@@ -1861,13 +1861,19 @@ std::string string_get_sortable_timestamp() {
void string_replace_all(std::string & s, const std::string & search, const std::string & replace) { void string_replace_all(std::string & s, const std::string & search, const std::string & replace) {
if (search.empty()) { if (search.empty()) {
return; // Avoid infinite loop if 'search' is an empty string return;
} }
std::string builder;
builder.reserve(s.length());
size_t pos = 0; size_t pos = 0;
while ((pos = s.find(search, pos)) != std::string::npos) { size_t last_pos = 0;
s.replace(pos, search.length(), replace); while ((pos = s.find(search, last_pos)) != std::string::npos) {
pos += replace.length(); builder.append(s, last_pos, pos - last_pos);
builder.append(replace);
last_pos = pos + search.length();
} }
builder.append(s, last_pos, std::string::npos);
s = std::move(builder);
} }
void string_process_escapes(std::string & input) { void string_process_escapes(std::string & input) {

14
examples/llava/clip.cpp
View File

@@ -216,13 +216,19 @@ static std::string gguf_data_to_str(enum gguf_type type, const void * data, int
static void replace_all(std::string & s, const std::string & search, const std::string & replace) { static void replace_all(std::string & s, const std::string & search, const std::string & replace) {
if (search.empty()) { if (search.empty()) {
return; // Avoid infinite loop if 'search' is an empty string return;
} }
std::string builder;
builder.reserve(s.length());
size_t pos = 0; size_t pos = 0;
while ((pos = s.find(search, pos)) != std::string::npos) { size_t last_pos = 0;
s.replace(pos, search.length(), replace); while ((pos = s.find(search, last_pos)) != std::string::npos) {
pos += replace.length(); builder.append(s, last_pos, pos - last_pos);
builder.append(replace);
last_pos = pos + search.length();
} }
builder.append(s, last_pos, std::string::npos);
s = std::move(builder);
} }
static std::string gguf_kv_to_str(const struct gguf_context * ctx_gguf, int i) { static std::string gguf_kv_to_str(const struct gguf_context * ctx_gguf, int i) {

14
src/llama-impl.h
View File

@@ -31,11 +31,17 @@ void llama_log_callback_default(ggml_log_level level, const char * text, void *
static void replace_all(std::string & s, const std::string & search, const std::string & replace) { static void replace_all(std::string & s, const std::string & search, const std::string & replace) {
if (search.empty()) { if (search.empty()) {
return; // Avoid infinite loop if 'search' is an empty string return;
} }
std::string builder;
builder.reserve(s.length());
size_t pos = 0; size_t pos = 0;
while ((pos = s.find(search, pos)) != std::string::npos) { size_t last_pos = 0;
s.replace(pos, search.length(), replace); while ((pos = s.find(search, last_pos)) != std::string::npos) {
pos += replace.length(); builder.append(s, last_pos, pos - last_pos);
builder.append(replace);
last_pos = pos + search.length();
} }
builder.append(s, last_pos, std::string::npos);
s = std::move(builder);
} }