llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard

Published 20 hours ago • verified publisher icon ggml-org

Refactoring: Use STL in input processing and fix off-by-one bug

Open thomasantony opened this issue 1 year ago • 0 comments

There was an off-by-one bug in the line

if (embd.size() > params.n_batch) {
    break;
}

it would copy n_batch+1 elements rather than n_batch. This PR fixes that and also switches to using STL instead of a while loop

thomasantony avatar Mar 14 '23 04:03 thomasantony