minja: add indent filter to support command-r-plus's chat templates

common/minja.hpp

@@ -2452,6 +2452,24 @@ inline std::shared_ptr<Context> Context::builtins() {
     }
     return res;
   }));
+  globals.set("indent", simple_function("indent", { "text", "indent", "first" }, [](const std::shared_ptr<Context> &, Value & args) {
+    auto text = args.at("text").get<std::string>();
+    auto first = args.get<bool>("first", false);
+    std::string out;
+    std::string indent(args.get<int64_t>("indent", 0), ' ');
+    std::istringstream iss(text);
+    std::string line;
+    auto is_first = true;
+    while (std::getline(iss, line, '\n')) {
+      auto needs_indent = !is_first || first;
+      if (is_first) is_first = false;
+      else out += "\n";
+      if (needs_indent) out += indent;
+      out += line;
+    }
+    if (!text.empty() && text.back() == '\n') out += "\n";
+    return out;
+  }));
   globals.set("selectattr", Value::callable([=](const std::shared_ptr<Context> & context, Value::Arguments & args) {
     args.expectArgs("selectattr", {2, std::numeric_limits<size_t>::max()}, {0, 0});
     auto & items = args.args[0];

tests/chat/contexts/tool_use.json

@@ -33,7 +33,7 @@
     },
     {
       "role": "assistant",
-      "content": null,
+      "content": "",
       "tool_calls": [
         {
           "id": "call_2",
@@ -60,7 +60,7 @@
     },
     {
       "role": "assistant",
-      "content": null,
+      "content": "",
       "tool_calls": [
         {
           "id": "call_3",

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-default-simple.txt

@@ -0,0 +1 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-default-system.txt

@@ -0,0 +1 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>You only tell the truth.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-rag-simple.txt

@@ -0,0 +1,16 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+## Task and Context
+You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
+
+## Style Guide
+Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results></results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
+Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
+Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
+Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-rag-system.txt

@@ -0,0 +1,12 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+You only tell the truth.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results></results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
+Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
+Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
+Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-rag-tool_use.txt

@@ -0,0 +1,16 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+## Task and Context
+You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
+
+## Style Guide
+Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Print a hello world message with python.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>Anything else?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Test a tautology.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>Truth is definitely true.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Check it on the web.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>I don't need the web to answer you but I did check, as you asked. What now?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results></results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
+Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
+Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
+Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-tool_use-simple.txt

@@ -0,0 +1,25 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+## Task and Context
+You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
+
+## Style Guide
+Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
+
+## Available Tools
+Here is a list of tools that you have available to you:
+
+<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
+```json
+[
+    {
+        "tool_name": title of the tool in the specification,
+        "parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
+    }
+]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-tool_use-system.txt

@@ -0,0 +1,21 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+You only tell the truth.
+
+## Available Tools
+Here is a list of tools that you have available to you:
+
+<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>What's your favourite LLM framework?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>llama.cpp!<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
+```json
+[
+    {
+        "tool_name": title of the tool in the specification,
+        "parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
+    }
+]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/goldens/CohereForAI-c4ai-command-r-plus-tool_use-tool_use.txt

@@ -0,0 +1,93 @@
+<|startoftext|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
+The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
+
+# System Preamble
+## Basic Rules
+You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
+
+# User Preamble
+## Task and Context
+You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
+
+## Style Guide
+Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
+
+## Available Tools
+Here is a list of tools that you have available to you:
+
+```python
+def ipython(code: str) -> List[Dict]:
+    """Runs code in an ipython interpreter and returns the result of the execution after 60 seconds.
+
+    Args:
+        code (str): The code to run in the ipython interpreter.
+    """
+    pass
+```
+
+```python
+def brave_search(query: str) -> List[Dict]:
+    """Executes a web search with Brave.
+
+    Args:
+        query (str): The query to search for.
+    """
+    pass
+```
+
+```python
+def wolfram_alpha(query: str) -> List[Dict]:
+    """Executes a query with Wolfram Alpha.
+
+    Args:
+        query (str): The query to execute.
+    """
+    pass
+```
+
+```python
+def test(condition: bool) -> List[Dict]:
+    """Runs a test.
+
+    Args:
+        condition (bool): The condition to test.
+    """
+    pass
+```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Print a hello world message with python.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
+Action:
+```json
+[
+    {
+        "tool_name": "ipython",
+        "parameters": "{\"code\": \"print('Hello, World!')\"}"
+    }
+]```
+<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
+{"stdout": "Hello, World!"}</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>Anything else?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Test a tautology.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
+Action:
+```json
+[
+    {
+        "tool_name": "test",
+        "parameters": "{\"condition\":true}"
+    }
+]```
+<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
+true</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>Truth is definitely true.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Check it on the web.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
+Action:
+```json
+[
+    {
+        "tool_name": "brave_search",
+        "parameters": "{\"query\": \"what is truth anyway am I right?\"}"
+    }
+]```
+<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
+{"title":"Truth: don't ask the web, ask an LLM instead!","url":"https://en.wikipedia.org/wiki/Truth"}</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>I don't need the web to answer you but I did check, as you asked. What now?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
+```json
+[
+    {
+        "tool_name": title of the tool in the specification,
+        "parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
+    }
+]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>

tests/chat/templates/CohereForAI-c4ai-command-r-plus-default.jinja

@@ -0,0 +1 @@
+{{ bos_token }}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% elif false == true %}{% set loop_messages = messages %}{% set system_message = 'You are Command-R, a brilliant, sophisticated, AI-assistant trained to assist human users by providing thorough responses. You are trained by Cohere.' %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% if system_message != false %}{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + system_message + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% elif message['role'] == 'assistant' %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'  + content.strip() + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}{% endif %}