Pause sampling if waiting for user input.

main.cpp

@@ -935,6 +935,7 @@ int main(int argc, char ** argv) {
         embd.clear();
 
         if (embd_inp.size() <= input_consumed) {
+            if (!is_interacting) {
                 // out of user input, sample next token
                 const float top_k = params.top_k;
                 const float top_p = params.top_p;
@@ -964,6 +965,7 @@ int main(int argc, char ** argv) {
 
                 // decrement remaining sampling budget
                 --remaining_tokens;
+            }
         } else {
             // some user input remains from prompt or interaction, forward it to processing
             while (embd_inp.size() > input_consumed) {