A 100‑sample Japanese instruction‑tuning evaluation dataset of annotated tasks—ranging from summarization correction and math reasoning to translation, creative generation, and user‑intent understanding—designed for manual or automatic 5‑point rating of fine‑tuned models.
Description in other languages
Statistics
Statistics as of Nov 16, 2026
Hugging Face downloads (last 30 days)
2,000
Hugging Face likes
103
Listed
—
Listed since the page was created
Downloads (30-day window) over time
Show as a table
| Month | DL (30d) |
|---|---|
| Dec 2025 | 4,000 |
| Jan 2026 | 3,000 |
| Feb 2026 | 2,000 |
| Mar 2026 | 2,000 |
| Apr 2026 | 2,000 |
| May 2026 | 3,000 |
| Jun 2026 | 2,000 |
| Jul 2026 | 2,000 |
| Aug 2026 | 2,000 |
| Sep 2026 | 2,000 |
| Oct 2026 | 2,000 |
| Nov 2026 | 2,000 |
Downloads over the 30 days before the last snapshot of each month.
Popular in the same category
- emilia-yodasTTS-AGI18kDataset of dialogues and lore from the Fate/Stay Night character “Emilia”, formatted for training and evaluating conversational language models.
- fgo_voices_jpdeepghs10kJP Voice‑Text Dataset for FGO Waifus: a 30,800‑record, 66.4‑hour Japanese audio collection of single‑voice‑actor character lines (≈7.76 s each) useful for ASR/ASV fine‑tuning and evaluation.
- JGLUEshunk0313kUpdated JGLUE dataset card and loading script for a Japanese NLP benchmark (created by Yahoo Japan and Waseda University) that covers text classification (MARC‑ja, JCoLA), sentence‑pair classification (JNLI), and QA (JSQuAD, JCommonsenseQA), with releases linked on GitHub and Hugging Face.
- JMTEBsbintuitions3kJMTEB is a Japanese text‑embedding benchmark featuring 5 tasks (clustering, classification, STS, retrieval, reranking) and 28 datasets, offering a one‑line evaluation script and inviting community contributions.
- JMMMUJMMMU3kJMMMU is a Japanese multimodal benchmark expanded over tenfold to 1,320 culturally diverse questions (720 culture‑agnostic, 600 culture‑specific) translated by native subject experts, now featuring a public leaderboard.
- JamC-QAsbintuitions2kJamC‑QA is a bilingual benchmark of multiple‑choice questions spanning eight Japanese‑culture and knowledge categories, with leaderboard metrics comparing state‑of‑the‑art models.
Resources in similar research fields
- qg_jaquadlmqg494Japanese JaQuAD, a subset of QG‑Bench, provides sentence‑ and paragraph‑level data with highlighted answer tokens for training Japanese question‑generation models, evaluated by BLEU4, METEOR, ROUGE‑L, BERTScore, and MoverScore.Shared: Annotation and Dataset Development
- zenz-v2.5-datasetMiwa-Keita439A 190 M‑pair JSONL dataset for Japanese kana‑to‑kanji conversion, featuring “left context – input – output” triples and pre‑trained models (medium, small, xsmall) plus the AJIMEE-Bench benchmark, released under CC BY‑SA 4.0 with subsets licensed from Wikipedia and Common Crawl.Shared: Annotation and Dataset Development
ajimee-benchazookey30AJIMEE-Bench (Advanced Japanese IME Evaluation Benchmark)GitHubShared: Annotation and Dataset Development- WildGuardTestJPsbintuitions258WildGuardTestJP is a Japanese‑language evaluation dataset of 1,725 samples, faithfully translated from WildGuardTest through a multi‑stage refinement pipeline (Seed‑X‑PPO‑7B, gpt‑oss‑120b, Qwen2.5‑72B‑Instruct, gemma‑3‑27b‑it) and released on Hugging Face under the ODC‑BY license.Shared: Annotation and Dataset Development
- bbh-japfnet213Japanese‑language BB‑Bench Hard dataset with evaluation questions (JSON Lines) and Chain‑of‑Thought prompts (YAML), translated via the PLaMo model.Shared: Annotation and Dataset Development
- covid_tweets_japanesecommunity-datasets171COVID‑19 Japanese Twitter dataset supplies Japanese tweet IDs and assessment codes (63–68) indicating COVID‑19 relevance and fact/opinion status, enabling text‑classification research.Shared: Annotation and Dataset Development