Support special tokens and not adding BOS to prompt in speculative

Author: AutonomicPerfectionist

Diff for: examples/speculative/speculative.cpp

@@ -94,9 +94,13 @@ int main(int argc, char ** argv) {
         }
     }
 
-    // tokenize the prompt
+
+    // Tokenize the prompt
+    const bool add_bos = llama_vocab_type(llama_get_model(ctx_tgt)) == LLAMA_VOCAB_TYPE_SPM;
+    LOG("add_bos: %d\n", add_bos);
+
     std::vector<llama_token> inp;
-    inp = ::llama_tokenize(ctx_tgt, params.prompt, true);
+    inp = ::llama_tokenize(ctx_tgt, params.prompt, add_bos, true);
 
     const int max_context_size     = llama_n_ctx(ctx_tgt);
     const int max_tokens_list_size = max_context_size - 4;