wip antislop

expose.h

@@ -3,6 +3,7 @@
 
 const int stop_token_max = 24;
 const int ban_token_max = 16;
+const int ban_phrase_max = 16;
 const int tensor_split_max = 16;
 const int logit_bias_max = 24;
 const int dry_seq_break_max = 24;
@@ -106,6 +107,7 @@ struct generation_inputs
     const float smoothing_factor = 0.0f;
     const logit_bias logit_biases[logit_bias_max] = {};
     const char * banned_tokens[ban_token_max] = {};
+    const char * banned_phrases[ban_phrase_max] = {};
 };
 struct generation_outputs
 {

gpttype_adapter.cpp

@@ -108,6 +108,7 @@ static std::vector<std::string> stop_sequence;
 static std::vector<int> special_stop_sequence; //for stop sequences that don't have a string representation
 static std::vector<std::string> banned_tokens;
 static std::vector<int> banned_token_ids;
+static std::vector<std::string> banned_phrases;
 static std::unordered_multimap<gpt_vocab::id, std::vector<gpt_vocab::id>> dry_sequence_breakers; // Multi-mapping from first token of sequence to tail of sequence (tail is empty for a single token)
 static std::vector<int> dry_repeat_count; // Indexed as last_n_tokens
 static std::unordered_map<gpt_vocab::id, int> dry_max_token_repeat;
@@ -2530,6 +2531,30 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
         }
     }
 
+    //antislop phrase banning
+    banned_phrases.clear();
+    delayed_generated_tokens_limit = 0;
+    for(int x=0;x<ban_phrase_max;++x)
+    {
+        std::string word = inputs.banned_phrases[x];
+        if(word!="")
+        {
+            std::vector<int> toks;
+            TokenizeString(word, toks, file_format, false);
+            int tokcount = toks.size();
+            if(tokcount>0)
+            {
+                tokcount += 2; //add some extra buffer
+            }
+            delayed_generated_tokens_limit = (tokcount>delayed_generated_tokens_limit?tokcount:delayed_generated_tokens_limit);
+            banned_phrases.push_back(word);
+        }
+    }
+    if(debugmode==1 && banned_phrases.size()>0)
+    {
+        printf("\nBanned a total of %zu phrases, with max token count of %d.\n",banned_phrases.size(),delayed_generated_tokens_limit);
+    }
+
     logit_biases.clear();
     for(int x=0;x<logit_bias_max;++x)
     {
@@ -3234,6 +3259,25 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
                 printf("]\n");
             }
 
+            //anti slop detection
+            // for (const auto &matched : stop_sequence)
+            // {
+            //     if (concat_output.find(matched) != std::string::npos)
+            //     {
+            //         stopper_unused_tokens = remaining_tokens;
+            //         remaining_tokens = 0;
+            //         if(allow_regular_prints)
+            //         {
+            //             auto match_clean = matched;
+            //             replace_all(match_clean, "\n", "\\n");
+            //             printf("\n(Stop sequence triggered: %s)", match_clean.c_str());
+            //         }
+            //         last_stop_reason = stop_reason::CUSTOM_STOPPER;
+            //         earlystopped = true;
+            //         break;
+            //     }
+            // }
+
             bool earlystopped = false;
             if(!inputs.bypass_eos_token && inputs.allow_eos_token && (id==eosID || (id==eotID && id!=-1)))
             {

koboldcpp.py

@@ -21,6 +21,7 @@ from datetime import datetime, timezone
 sampler_order_max = 7
 stop_token_max = 24
 ban_token_max = 16
+ban_phrase_max = 16
 tensor_split_max = 16
 logit_bias_max = 24
 dry_seq_break_max = 24
@@ -171,7 +172,8 @@ class generation_inputs(ctypes.Structure):
                 ("dynatemp_exponent", ctypes.c_float),
                 ("smoothing_factor", ctypes.c_float),
                 ("logit_biases", logit_bias * logit_bias_max),
-                ("banned_tokens", ctypes.c_char_p * ban_token_max)]
+                ("banned_tokens", ctypes.c_char_p * ban_token_max),
+                ("banned_phrases", ctypes.c_char_p * ban_phrase_max)]
 
 class generation_outputs(ctypes.Structure):
     _fields_ = [("status", ctypes.c_int),
@@ -910,6 +912,7 @@ def generate(genparams, is_quiet=False, stream_flag=False):
     logit_biases = genparams.get('logit_bias', {})
     render_special = genparams.get('render_special', False)
     banned_tokens = genparams.get('banned_tokens', [])
+    banned_phrases = genparams.get('banned_phrases', [])
     bypass_eos_token = genparams.get('bypass_eos', False)
 
     inputs = generation_inputs()
@@ -1028,6 +1031,12 @@ def generate(genparams, is_quiet=False, stream_flag=False):
         else:
             inputs.banned_tokens[n] = banned_tokens[n].encode("UTF-8")
 
+    for n in range(ban_phrase_max):
+        if not banned_phrases or n >= len(banned_phrases):
+            inputs.banned_phrases[n] = "".encode("UTF-8")
+        else:
+            inputs.banned_phrases[n] = banned_phrases[n].encode("UTF-8")
+
     currentusergenkey = genkey
     totalgens += 1
     #early exit if aborted