Sign inGet started
Dataset · Hugging Face

MC7ever/merged-text-3split

merged-text-3split Concatenation of every local dataset and remote Hub dataset in the merge manifest, normalised to a single text-only tabular schema and split three ways. Splits split rows shards train 478,621 10 bench 47,940 1 eval 12,270 1 Held-out eval and bench are deterministic (sha1 of row id), so re-running the merge reproduces byte-identical assignments. Schema (all columns are text) id source source_split domain format… See the full description on the dataset page: https://huggingface.co/datasets/MC7ever/merged-text-3split.

View on Hugging Face
Report an issue
Author / organizationMC7ever
License (as declared)other
Downloads0 (as reported by Hugging Face)
Likes0
Created18 October 2026
Last updated at source18 October 2026
roleplayerpcreative-writingeroticansfwsftinstruction-tuningconversationalchatpersonastorytellingreasoningmathcodeagenttool-usedpopreferencebenchmarkevaltext-to-textllm-trainingnot-for-all-audiences
Embed this entry
<iframe src="https://silytics.si/embed/item/cf65e37f-d353-4007-8ab3-1689767e36e7" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Recent changes

  • 18 Oct 2026 Topics added roleplay, erp, creative-writing, erotica, nsfw, sft, instruction-tuning, conversational, chat, persona, storytelling, reasoning, math, code, agent, tool-use, dpo, preference, benchmark, eval

Changes detected by Silytics between refreshes of the source.

More from MC7ever

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.