h2o-llmstudio icon indicating copy to clipboard operation
h2o-llmstudio copied to clipboard

[FEATURE] Add ability to specify dataset problem type during data import

Open MartinBarus opened this issue 9 months ago • 1 comments

🚀 Feature

LLM studio supports multiple problem types (Causal Modeling/ Classification), however, during data import, it expects dataset to be named exactly as the problem type, otherwise proper problem type can not be set.

Motivation

When importing custom dataset named train.csv for classification, it would be great to have the ability to select proper problem type.

MartinBarus avatar Apr 26 '24 19:04 MartinBarus

Maybe add this snippet

items += [
                ui.dropdown(
                    name="dataset/import/cfg_file",
                    label="Problem Type",
                    required=True,
                    choices=choices_problem_types,
                    value=q.client["dataset/import/cfg_file"],
                    trigger=True,
                    tooltip=tooltips["experiments_problem_type"],
                )
            ]

here :https://github.com/h2oai/h2o-llmstudio/blob/d1acb59/llm_studio/app_utils/sections/dataset.py#L491

MartinBarus avatar Apr 26 '24 19:04 MartinBarus