disable debug prints

Author: ddh0

src/llama-memory-recurrent.cpp

@@ -136,7 +136,7 @@ void llama_memory_recurrent::clear(bool data) {
 }
 
 bool llama_memory_recurrent::seq_rm(llama_seq_id seq_id, llama_pos p0, llama_pos p1) {
-    printf("[DEBUG] calling llama_memory_recurrent::seq_rm` with `seq_id=%d, p0=%d, p1=%d`\n", seq_id, p0, p1);
+    //printf("[DEBUG] calling llama_memory_recurrent::seq_rm` with `seq_id=%d, p0=%d, p1=%d`\n", seq_id, p0, p1);
     uint32_t new_head = size;
 
     if (p0 < 0) {
@@ -158,7 +158,7 @@ bool llama_memory_recurrent::seq_rm(llama_seq_id seq_id, llama_pos p0, llama_pos
             const auto & cell = cells[tail_id];
             // partial intersection is invalid
             if ((0 < p0 && p0 < cell.pos) || (0 < p1 && p1 <= cell.pos)) {
-                printf("[DEBUG] inside `llama_memory_recurrent::seq_rm`: partial intersection is invalid, so returning false\n");
+                //printf("[DEBUG] inside `llama_memory_recurrent::seq_rm`: partial intersection is invalid, so returning false\n");
                 return false;
             }
             // invalidate tails which will be cleared
@@ -169,7 +169,7 @@ bool llama_memory_recurrent::seq_rm(llama_seq_id seq_id, llama_pos p0, llama_pos
     } else {
         // seq_id is negative, then the range should include everything or nothing
         if (p0 != p1 && (p0 != 0 || p1 != std::numeric_limits<llama_pos>::max())) {
-            printf("[DEBUG] inside `llama_memory_recurrent::seq_rm`: `seq_id` is negative, so returning false\n");
+            //printf("[DEBUG] inside `llama_memory_recurrent::seq_rm`: `seq_id` is negative, so returning false\n");
             return false;
         }
     }

tools/server/server.cpp

@@ -3565,7 +3565,7 @@ struct server_context {
                                     );
 
                                     bool do_reset = it == slot.ctx_checkpoints.rend();
-                                    printf("[DEBUG] `do_reset` was set to `%s`\n", do_reset ? "true" : "false");
+                                    //printf("[DEBUG] `do_reset` was set to `%s`\n", do_reset ? "true" : "false");
 
                                     if (!do_reset) {
                                         // restore the context checkpoint
@@ -3575,7 +3575,7 @@ struct server_context {
                                         if (n != ctx_checkpoint_size) {
                                             SLT_ERR(slot, "failed to restore context checkpoint (pos_min = %d, pos_max = %d, size = %.3f MiB)\n", it->pos_min, it->pos_max, (float) ctx_checkpoint_size / 1024 / 1024);
                                             do_reset = true;
-                                            printf("[DEBUG] `do_reset` was set to `true` after failing to restore a checkpoint");
+                                            //printf("[DEBUG] `do_reset` was set to `true` after failing to restore a checkpoint");
                                         } else {
                                             slot.n_past = std::min(slot.n_past, it->pos_max);
                                             SLT_WRN(slot, "restored context checkpoint (pos_min = %d, pos_max = %d, size = %.3f MiB)\n", it->pos_min, it->pos_max, (float) ctx_checkpoint_size / 1024 / 1024);
@@ -3630,7 +3630,6 @@ struct server_context {
                         // there is no common part left
                         slot.n_past                = 0;
                         slot.n_prompt_tokens_cache = 0;
-                        printf("[DEBUG] we had no choice but to truncate all tokens from this slot :( very sad");
                     }
 
                     SLT_INF(slot, "n_past = %d\n", slot.n_past);