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

minor changes

Browse source
This commit is contained in:
pudepiedj 2024-02-19 12:41:27 +00:00
parent 7f0d8987eb
commit 8a4d202957
2 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

2
Llamaserver.py
View file

@ -70,7 +70,7 @@ def send_request(q, question, event, count, num_requests):
bar = make_progress_bar(bar, count, num_requests)
q.task_done()
elif response.status_code == 429 and not q.empty():
event.set()
# event.set()
print("Server return too many requests; back off!! Reset event.")
else:
print(f"Server responded with code {response.status_code}\n")

4
examples/server/server.cpp
View file

@ -324,7 +324,7 @@ static void kvgraphics(std::vector<llama_client_slot>& slots, int cache_size) {
std::string slot_symbol3 = "";
#ifdef DEBUG
return; // do not display graphics when in debug build
return; // do not display graphics when in debug build (doesn't seem to work)
#endif
// return if empty
@ -374,7 +374,7 @@ static void kvgraphics(std::vector<llama_client_slot>& slots, int cache_size) {
}
printf(" %4zu/%5zu %2d %s %s %s\n", slots[i].cache_tokens.size(), slot_cache_size, slots[i].id, slot_symbol1.c_str(), slot_symbol2.c_str(), slot_symbol3.c_str());
}
printf("\n\033[%dJ", 0);
printf("\n\033[%dJ", num_blocks); // move cursor to end of cache display
}
struct llama_server_context