Got Hq Merged 75K: High-Quality Multi-Source Text Dataset

WithinUsAI created a dataset of 75,000 text examples on 2026-05-30. It merges content from 16 diverse sources by sampling 5,000 examples from each. The dataset includes a category for reasoning, which comprises 25,000 examples.

Use Cases

• Training language models on high-quality text based on the description of a normalized, multi-source collection.
• Benchmarking model performance on reasoning tasks based on the 'reasoning' category mentioned.
• Studying dataset merging and normalization techniques based on the described unified schema process.
• Evaluating model generalization across diverse sources based on the 16-source sampling strategy.

Strengths

• Contains 75,000 total examples.
• Integrates data from 16 distinct sources.
• Includes 25,000 examples categorized as 'reasoning'.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• The specific sources and sampling methodology beyond the count are not detailed.
• Freshness should be verified as the last update was 2026-05-30.

Provenance

Source

WithinUsAI

Collection Method

Sampled 5,000 examples from each of 16 sources, normalized into a unified schema.

Freshness

Last updated 2026-05-30 08:44:16.