Revert "server : support multi-modal context checkpoints and prompt caching (ikawrakow#1398)"

This reverts commit 433531d.

Author: Nexesenex

common/common.h

@@ -422,7 +422,7 @@ struct gpt_params {
     float slot_prompt_similarity = 0.1f;
 
     bool do_checkpoint = false;               // do checkpoint for recurrent models only
-    int32_t ctx_checkpoints_n = 32;           // max number of context checkpoints per slot
+    int32_t ctx_checkpoints_n = 8;            // max number of context checkpoints per slot
     int32_t ctx_checkpoints_interval = 512;   // minimum number of tokens between each context checkpoints
     int32_t ctx_checkpoints_tolerance = 5;    // the number of tokens before the full prompt to create the checkpoint 
     int32_t cache_ram_mib = 8192;   // -1 = no limit, 0 - disable, 1 = 1 MiB, etc.

common/regex-partial.cpp

@@ -101,7 +101,7 @@ std::string regex_to_reversed_partial_regex(const std::string & pattern) {
                 sequence->back() += *it;
                 auto is_star = *it == '*';
                 ++it;
-                if (it != end && is_star) {
+                if (is_star) {
                     if (*it == '?') {
                         ++it;
                     }