[BUG-python/deployment] Logging several dataset subsets in the same argilla dataset override records

[frascuchon]

Describe the bug

Loading an HF dataset with several subsets into a single Argilla dataset overrides the records when the dataset has no row id. Subsets have the same structure, so we could load them into the same Argilla dataset, but users need to build the record.id attribute to avoid overrides.

Stacktrace and Code to create the bug

dataset = rg.Dataset(...)

for config in ["gpqa_diamond", "gpqa_extended", "gpqa_main"]:
    hf_ds = load_dataset("some-dataset-with-several-subsets", name=config, token=HF_TOKEN, split="train")
    dataset.records.log(hf_ds.map(lambda r: {"subset": config})) # row_id + split will be use to identify record id -> overriding records

Expected behavior

Environment:

• Argilla Version [e.g. 1.0.0]:
• ElasticSearch Version [e.g. 7.10.2]:
• Docker Image (optional) [e.g. argilla:v1.0.0]:

Additional context