DeepSeek R1: parse thoughts / return in separate field in API (non streamed mode)

common/chat.cpp

@@ -22,6 +22,18 @@ std::string common_chat_format_name(common_chat_format format) {
     }
 }
 
+static std::string string_trim(const std::string & s) {
+    size_t start = 0;
+    while (start < s.size() && std::isspace(s[start])) {
+        start++;
+    }
+    size_t end = s.size();
+    while (end > start && std::isspace(s[end - 1])) {
+        end--;
+    }
+    return s.substr(start, end - start);
+}
+
 const common_grammar_options grammar_options {
     /* .dotall = */ false,
     /* .compact_spaces = */ false,
@@ -537,20 +549,43 @@ static common_chat_params common_chat_params_init_deepseek_r1(const common_chat_
         });
         data.grammar_triggers.push_back({"<｜tool▁calls▁begin｜>", /* .at_start = */ false});
         data.preserved_tokens = {
+            "<think>",
+            "</think>",
             "<｜tool▁sep｜>",
             "<｜tool▁call▁end｜>",
         };
         builder.add_rule("root", "\"<｜tool▁calls▁begin｜>\" (" + string_join(tool_rules, " | ") + ")" + (inputs.parallel_tool_calls ? "*" : "") + " space");
     }, grammar_options);
-    data.prompt = tmpl.apply(inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt);
+    /*
+        Note: we do not feed the thoughts back to the template for a few reasons:
+        - the template doesn't use them explicitly
+        - if content isn't null, tool calls arent rendered
+        - not having the thoughts will locally reset the KV cache (losing the hot tokens of the tool calls) but will save up a lot long term.
+    */
+    auto prompt = tmpl.apply(inputs.messages, inputs.tools.empty() ? json() : inputs.tools, inputs.add_generation_prompt);
+    std::string suffix = "<｜Assistant｜>";
+    if (vocab && !llama_vocab_get_add_eos(vocab) &&
+        inputs.add_generation_prompt &&
+        !string_ends_with(prompt, suffix))
+    {
+        prompt += "<｜end▁of▁sentence｜>";
+    }
+    data.prompt = prompt;
     data.format = COMMON_CHAT_FORMAT_DEEPSEEK_R1;
     return data;
 }
 static common_chat_msg common_chat_parse_deepseek_r1(const std::string & input) {
     static std::regex trigger_regex("<｜tool▁calls▁begin｜>");
-    static std::regex function_regex("<｜tool▁call▁begin｜>function<｜tool▁sep｜>([^\n]+)\n```json\n");
+    static std::regex function_regex(R"(<｜tool▁call▁begin｜>function<｜tool▁sep｜>([^\n]+)\n```json\n)");
     static std::regex close_regex("```<｜tool▁call▁end｜>");
-    return parse_json_tool_calls(input, trigger_regex, function_regex, close_regex);
+    static std::regex think_regex(R"(<think>([\s\S\n]*)</think>([\s\S\r\n]*))");
+    auto msg = parse_json_tool_calls(input, trigger_regex, function_regex, close_regex);
+    std::smatch match;
+    if (std::regex_match(msg.content, match, think_regex)) {
+        msg.thoughts = string_trim(match[1].str());
+        msg.content = string_trim(match[2].str());
+    }
+    return msg;
 }
 
 static common_chat_params common_chat_params_init_firefunction_v2(const common_chat_template & tmpl, const struct common_chat_inputs & inputs) {

common/common.h

@@ -623,6 +623,7 @@ struct common_chat_msg {
     std::string role;
     std::string content;
     std::vector<common_tool_call> tool_calls;
+    std::string thoughts = "";
     std::string tool_plan = "";
 };
 

examples/server/server.cpp

@@ -744,6 +744,9 @@ struct server_task_result_cmpl_final : server_task_result {
             {"tool_calls", tool_calls},
             {"role", "assistant"},
         };
+        if (!msg.thoughts.empty()) {
+            message["thoughts"] = msg.thoughts;
+        }
         if (!msg.tool_plan.empty()) {
             message["tool_plan"] = msg.tool_plan;
         }

tests/test-chat.cpp

@@ -18,18 +18,17 @@
 using json = nlohmann::ordered_json;
 
 static common_chat_msg msg_from_json(const json & message) {
-    common_chat_msg ret{
+    common_chat_msg ret;
-        "assistant",
+    ret.role = "assistant";
-        "",
-        {},
-        /* .tool_plan = */ "",
-    };
     if (message.contains("content") && !message.at("content").is_null()) {
         ret.content = message.at("content");
     }
     if (message.contains("tool_plan")) {
         ret.tool_plan = message.at("tool_plan");
     }
+    if (message.contains("thoughts")) {
+        ret.thoughts = message.at("thoughts");
+    }
     auto has_tool_calls = message.contains("tool_calls");
     if (has_tool_calls) {
         for (const auto & tc : message.at("tool_calls")) {