The text I generated using this repository is inconsistent with the official output.

[yanrui27]

I used the official code from https://cookbook.openai.com/articles/openai-harmony
, but the generated text is inconsistent with the one shown in the cookbook.

The main issue occurs when calling tools.

In the official example, the format is:

<|start|>assistant<|channel|>commentary to=functions.get_current_weather <|constrain|>json<|message|>

But when I use the code provided in the repository, the generated result is:

<|start|>assistant to=functions.get_current_weather<|channel|>commentary <|constrain|> json<|message|>

The position of to= is different between the two outputs.

In the official example, it appears after <|channel|>commentary, while in the generated output from the provided code, it appears immediately after <|start|>assistant.

---

from openai_harmony import (
    Author,
    Conversation,
    DeveloperContent,
    HarmonyEncodingName,
    Message,
    Role,
    SystemContent,
    ToolDescription,
    load_harmony_encoding,
    ReasoningEffort
)
 
encoding = load_harmony_encoding(HarmonyEncodingName.HARMONY_GPT_OSS)
 
system_message = (
    SystemContent.new()
        .with_reasoning_effort(ReasoningEffort.HIGH)
        .with_conversation_start_date("2025-06-28")
)
 
developer_message = (
    DeveloperContent.new()
        .with_instructions("Always respond in riddles")
        .with_function_tools(
            [
                ToolDescription.new(
                    "get_current_weather",
                    "Gets the current weather in the provided location.",
                    parameters={
                        "type": "object",
                        "properties": {
                            "location": {
                                "type": "string",
                                "description": "The city and state, e.g. San Francisco, CA",
                            },
                            "format": {
                                "type": "string",
                                "enum": ["celsius", "fahrenheit"],
                                "default": "celsius",
                            },
                        },
                        "required": ["location"],
                    },
                ),
            ]
	)
)
 
convo = Conversation.from_messages(
    [
        Message.from_role_and_content(Role.SYSTEM, system_message),
        Message.from_role_and_content(Role.DEVELOPER, developer_message),
        Message.from_role_and_content(Role.USER, "What is the weather in Tokyo?"),
        Message.from_role_and_content(
            Role.ASSISTANT,
            'User asks: "What is the weather in Tokyo?" We need to use get_current_weather tool.',
        ).with_channel("analysis"),
        Message.from_role_and_content(Role.ASSISTANT, '{"location": "Tokyo"}')
        .with_channel("commentary")
        .with_recipient("functions.get_current_weather")
        .with_content_type("<|constrain|> json"),
        Message.from_author_and_content(
            Author.new(Role.TOOL, "functions.get_current_weather"),
            '{ "temperature": 20, "sunny": true }',
        ).with_channel("commentary"),
    ]
)
 
 
tokens = encoding.render_conversation_for_completion(convo, Role.ASSISTANT)
 
print(encoding.decode(tokens))

[levunet]

This issue has been resolved in this PR. Since it hasn't been reviewed yet, I'd recommend using it separately if you need it. #76