Skill trainer FAQ
    • PDF

    Skill trainer FAQ

    • PDF

    Article Summary

    The latest service changes have not yet been reflected in this content. We will update the content as soon as possible. Please refer to the Korean version for information on the latest updates.

    Available in Classic and VPC

    The following describes frequently occurring problems and how to troubleshoot them.

    Scenario

    Q. If I edit a value in the skill data area, the next step disappears. How can I load it?
    A. Edit the value in the skill data area and click the [Apply] button. It is normal that when you edit your skill data values (thoughts, actions, action input), observation values in the skill data area and values in the planning data area are initialized. When you click the [Apply] button, the created result is reloaded and the next step is recreated.

    Q. In the process of converting a user query into an action input value (URL), the language model does not properly generate parameters. How do I set this up to generate action input values as intended?
    A. Add an example to the description of the parameters in API Spec. With examples, the language model can understand user queries based on API Spec and generate URLs with correct parameters.
    <writing example>

    {
        "keyword": {
            "type": "string",
            "description": "a modifier in the form of an adjective or an adverb. \<example> kind, satisfactory, good",
        },
        "review": {
            "type": "string",
            "description": "positive or negative assessment. \<example> either pos or neg"
        }
    }
    

    Q. I created a skill and tried to collect scenarios, but the notification says it's impossible. How can I solve it?
    A. Check if the skill status says "completed."
    For the skill to be "completed," the following conditions should be met.

    • Fill in all required information
    • Authenticate API Spec
    • Check for duplicates in skill names

    Q. An error has occurred during a scenario task. How can I check the log?
    A. If an error occurs during a task, send us an inquiry with the following information. We will walk you through the troubleshooting steps.

    • Error screen (screen shot)
    • Skill set
    • Skill information

    Q. I entered a user query and clicked the [Run] button, but the planning data was not created. How can I resolve this?
    A. If the model does not properly understand the skill's API Spec or Manifest, planning data may not be generated. If you wrote the API Spec and Manifest information in English, please rewrite them in Korean and edit the skill information.

    Skill tuning

    Q. If I re-tune a tuned skill set, will the existing tuning model be deleted?
    A. The existing tuning model is not deleted, and an additional tuning model is created.

    Q. How many scenarios should I create when I tune my skill set? Tell me about the minimum and maximum number of scenarios that guarantee performance.
    A. We recommend creating 50 to 100 scenarios per skill set. The minimum number of scenarios for each tuning item is as follows: Depending on API complexity, the required number of scenarios may vary.

    Q. I tried to edit and tune the skill data area, but the tuning results have not been applied properly.
    A. To enhance the model's understanding, it is recommended to write a detailed description of the object targeted for tuning in the API Spec. Refer to the examples provided in Skill management for guidance. Note that to apply the edited API Spec, you will need to re-collect the scenarios.

    Q. I clicked the [Create task] button to tune the skill set, but the skill set I created is not listed.
    A. Check if the skill status says "completed." You can only tune a skill that is completed.

    Q. How many scenarios do I need to tune my skill set?
    A. You must have at least 1 scenario in "completed" status before you can tune your skill set. However, we recommend collecting more scenarios to improve skill set performance.

    Q. If I select a model tuned with 9 scenarios and re-tune 1 scenario, do I get a tuning model that reflects 10 scenarios?
    A. Yes. Since all scenarios in the "completed" status at the time of the tuning request are subject to tuning, you'll get a tuning model that reflects all 10 scenarios.

    Skill set

    Q. What is the maximum number of skills I can use in a multi-skill scenario?
    A. You can use up to 2 skills in a multi-skill scenario. If you use 3 skills or more, we can't guarantee the performance of the language model.

    Q. How many skills can I create in a skill set?
    A. There is no limit, but we recommend creating up to 4.

    Skill

    Q. Is there a designated port that I need to enter in my skill Endpoint URL?
    A. There is no designated port. But errors may occur due to communication issues or model performance issues related to Endpoint. If the problem persists, please send us an inquiry with a screenshot.

    API Spec

    Q. Can I write a description for API Spec in English?
    A. Yes, you can. But we recommend writing in Korean to maximize the performance of the model.

    Q. Does the Endpoint URL of the skill support only https?
    A. Endpoint URL supports both http and https.

    Q. Can I configure a URL in the path method for API calls?
    A. We recommend using the argument method for API URL configuration. Since all training data used in the skill model is based on the argument method, performance cannot be guaranteed if you configure the URL using the path method.

    Q. What should I write for a parameter description?
    A. Offer examples in a parameter description. If there are examples, the model can create a more accurate action input value based on the examples in the description.

    {
        "date": {
            "type": "string",
            "description": "base date. Must be in yyyymmdd format. \<example> 20230101. 20231121. 20210403"
        }
    }
    

    Manifest

    Q. How should I write the Description for model for my multi-path API to work properly?
    A. Description for the model is used as information to select an appropriate path for user queries when there are multiple paths within one skill. So, you need to specify the roles for each path and provide keywords to be searched for each path.
    The performance of the language model can be improved by specifying path information and parameter information separately as follows:

    • Beverage query skill: when searching for a beverage, use /sample_api/beverage. This API operates well when you enter a beverage name as a search keyword.
    • Liquor query skill: when searching for liquor, use /sample_api/liquor. This API operates well when you enter a liquor name as a search keyword.
    • Discounted item search skill: when searching for daily discounted items, use /sample/event. This API operates when you enter a date as a search keyword, and the search date must be entered in yyyymmdd format.

    Was this article helpful?

    What's Next
    Changing your password will log you out immediately. Use the new password to log back in.
    First name must have atleast 2 characters. Numbers and special characters are not allowed.
    Last name must have atleast 1 characters. Numbers and special characters are not allowed.
    Enter a valid email
    Enter a valid password
    Your profile has been successfully updated.