diff --git a/examples/common.cpp b/examples/common.cpp
index 53afe3cc2..9b8e2c39c 100644
--- a/examples/common.cpp
+++ b/examples/common.cpp
@@ -125,7 +125,7 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
             }
             params.path_prompt_cache = argv[i];
         } else if (arg == "--prompt-cache-all") {
-            params.prompt_cache_save_all = true;
+            params.prompt_cache_all = true;
         } else if (arg == "-f" || arg == "--file") {
             if (++i >= argc) {
                 invalid_param = true;
@@ -370,7 +370,8 @@ void gpt_print_usage(int /*argc*/, char ** argv, const gpt_params & params) {
     fprintf(stderr, "                        prompt to start generation with (default: empty)\n");
     fprintf(stderr, "  -e                    process prompt escapes sequences (\\n, \\r, \\t, \\', \\\", \\\\)\n");
     fprintf(stderr, "  --prompt-cache FNAME  file to cache prompt state for faster startup (default: none)\n");
-    fprintf(stderr, "  --prompt-cache-all    if specified, saves user input and generations to cache as well\n");
+    fprintf(stderr, "  --prompt-cache-all    if specified, saves user input and generations to cache as well.\n");
+    fprintf(stderr, "                        not supported with --interactive or other interactive options\n");
     fprintf(stderr, "  --random-prompt       start with a randomized prompt.\n");
     fprintf(stderr, "  --in-prefix STRING    string to prefix user inputs with (default: empty)\n");
     fprintf(stderr, "  --in-suffix STRING    string to suffix after user inputs with (default: empty)\n");
diff --git a/examples/common.h b/examples/common.h
index 00ddd7732..499671b2e 100644
--- a/examples/common.h
+++ b/examples/common.h
@@ -54,11 +54,11 @@ struct gpt_params {
     std::string lora_adapter = "";  // lora adapter path
     std::string lora_base = "";     // base model path for the lora adapter
 
-    bool memory_f16            = true;  // use f16 instead of f32 for memory kv
-    bool random_prompt         = false; // do not randomize prompt if none provided
-    bool use_color             = false; // use color to distinguish generations and inputs
-    bool interactive           = false; // interactive mode
-    bool prompt_cache_save_all = false; // save user input and generations to prompt cache
+    bool memory_f16        = true;  // use f16 instead of f32 for memory kv
+    bool random_prompt     = false; // do not randomize prompt if none provided
+    bool use_color         = false; // use color to distinguish generations and inputs
+    bool interactive       = false; // interactive mode
+    bool prompt_cache_all  = false; // save user input and generations to prompt cache
 
     bool embedding         = false; // get only sentence embedding
     bool interactive_first = false; // wait for user input immediately
diff --git a/examples/main/main.cpp b/examples/main/main.cpp
index bb172cce2..6604e60f2 100644
--- a/examples/main/main.cpp
+++ b/examples/main/main.cpp
@@ -140,7 +140,6 @@ int main(int argc, char ** argv) {
     params.prompt.insert(0, 1, ' ');
 
     std::string path_session    = params.path_prompt_cache;
-    const bool session_save_all = params.prompt_cache_save_all;
     std::vector<llama_token> session_tokens;
 
     if (!path_session.empty()) {
@@ -236,6 +235,11 @@ int main(int argc, char ** argv) {
     }
 
     if (params.interactive) {
+        if (params.prompt_cache_all) {
+            fprintf(stderr, "error: --prompt-cache-all not supported in interactive mode yet\n");
+            return 1;
+        }
+
 #if defined (__unix__) || (defined (__APPLE__) && defined (__MACH__))
         struct sigaction sigint_action;
         sigint_action.sa_handler = sigint_handler;
@@ -295,8 +299,7 @@ int main(int argc, char ** argv) {
 
     bool is_antiprompt        = false;
     bool input_echo           = true;
-    bool need_to_save_session = !path_session.empty();
-
+    bool need_to_save_session = !path_session.empty() && n_matching_session_tokens < embd_inp.size();
 
     int n_past             = 0;
     int n_remain           = params.n_predict;
@@ -325,11 +328,7 @@ int main(int argc, char ** argv) {
                 embd.insert(embd.begin(), last_n_tokens.begin() + n_ctx - n_left/2 - embd.size(), last_n_tokens.end() - embd.size());
 
                 // stop saving session if we run out of context
-                if (!path_session.empty() && session_save_all) {
-                    llama_save_session_file(ctx, path_session.c_str(),
-                        session_tokens.data(), session_tokens.size());
-                }
-                path_session = "";
+                path_session.clear();
 
                 //printf("\n---\n");
                 //printf("resetting: '");
@@ -604,7 +603,7 @@ int main(int argc, char ** argv) {
         }
     }
 
-    if (!path_session.empty() && session_save_all) {
+    if (!path_session.empty() && params.prompt_cache_all) {
         fprintf(stderr, "\n%s: saving final output to session file '%s'\n", __func__, path_session.c_str());
         llama_save_session_file(ctx, path_session.c_str(), session_tokens.data(), session_tokens.size());
     }