agentlans/m-a-p-FineFineWeb-sample

Unofficial m-a-p/FineFineWeb Sample This dataset is a processed, lightweight sample of the original m-a-p/FineFineWeb, a comprehensive corpus designed for fine-grained domain web text studies. Sampling Methodology To create this subset, the following processing steps were taken: Selection: 100 random .jsonl files were chosen from the original dataset. Extraction: 10,000 rows were downloaded per selected file. Processing: The extracted rows were combined and… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/m-a-p-FineFineWeb-sample.