Create your own evaluation datasets
The Custom Eval feature in H2O LLM DataStudio enables you to create your own evaluation datasets from various document formats (such as PDFs and DOC files), from audio and video files and existing datasets. These evaluation datasets can be downloaded in JSON formats for each evaluation type, allowing for seamless integration with H2O Eval Studio.
note
Custom Eval only supports English language.
Instructions
To create your own evaluation dataset, consider the following instructions:
- On the H2O LLM DataStudio left navigation menu, click Custom Eval.
- On the Create Your Own Eval Datasets page, click New.
- On the Project name text box, enter a name for the project.
- On the Description text box, enter a description for the project.
- On the Dataset type drop-down menu, select the evaluation dataset type. The available dataset types are,
- Question type: In this dataset, each entry includes a specific question, the correct answer to that question, and a label indicating the type of question (e.g., simple, conditional).
- Multi-Choice: In this dataset, each entry consists of a question followed by multiple answer choices, and the correct, or most appropriate answer.
- Token presence: In this dataset, each entry includes a question, the correct answer, and a list of key tokens that are relevant to the answer.
- On the Do you already have a QA Dataset? drop-down menu, select the following options accordingly:
- Yes (already possess a question-answer dataset)
- No (Do not have a pre-existing question-answer dataset)
Select Yes already possess a question-answer dataset.
7. Click Next.
8. Click Browse to upload the Q&A dataset in CSV file format.
9. Click Upload.
7. Click Next.
8. Click Browse to upload the Q&A dataset in CSV file format.
9. Click Upload.
If you do not have a pre-existing question-answer dataset, select No, and the LLM DataStudio will create question-answer pairs for you. It will generate these pairs by applying data curation techniques to the document you upload and subsequently use those question-answer pairs to generate an evaluation dataset.
Please note that generating an Eval Dataset from Curate may take a long time to complete.
7. Click Next.
8. Click Browse to upload the document or add the webpage URL if you are generating question-answer pairs from a webpage, or PDF web URL.
9. Click Upload.
Please note that generating an Eval Dataset from Curate may take a long time to complete.
7. Click Next.
8. Click Browse to upload the document or add the webpage URL if you are generating question-answer pairs from a webpage, or PDF web URL.
9. Click Upload.
- Under Configure columns section, select the columns which contain the context, question, and answer from the given options.
- Click Run pipeline.
Feedback
- Submit and view feedback for this page
- Send feedback about H2O LLM DataStudio | Docs to cloud-feedback@h2o.ai