Three-class Turkish sentiment analysis dataset
Corpus of 100 thousand to 1 million Turkish sentences labeled as positive, negative, and neutral.
- Text classification
- Turkish
Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.
Answered by Gemini 4 · the chat is kept only in this tab
Corpus of 100 thousand to 1 million Turkish sentences labeled as positive, negative, and neutral.
A cleaned English SFT dataset from Claude Fable-5 agent traces, 12,730 records.
Between 1 and 10 million semantically equivalent sentences from Tatoeba for 73 languages.
A set of 1k–10k challenging English text tasks where LLMs historically lagged behind humans.
Hundreds of thousands of structured character, artist, and general tags without the images themselves.
English feedback texts to train open LLMs in fine-grained long answer evaluation.
35k English patents and abstracts divided into 9 classes for long-context evaluation.
150,000 English examples for training models on choice selection, statement evaluation, and scoring without free text generation.
An English tabular and text corpus for sentiment classification and translation tasks.
A collection of research papers, code, chats, and traces from autonomous scientific agents on a preprint server.
Over 200k arXiv paper titles and abstracts labeled across 130 thematic classes.
Manually annotated snippets of real web texts in 109 languages for evaluating language recognition algorithms.
Multilingual compilation of datasets for fine-tuning embeddings for retrieval, reranking, similarity, and classification.
A massive English corpus of over 15.6 million articles from Fandom.com for pre-training and text generation.
A set of 248 corporate document images for training classifiers.
276k instruction-response pairs and multi-turn dialogues in French for language model training.
A corpus of scientific conference reviews with LLM-extracted claims and metrics.
Dataset for evaluating the safety of both final answers and intermediate reasoning steps of models.
Over 1.5 million ratings from real people for images generated from English text prompts.
A large corpus of English developer questions for training ticket priority classifiers.
English benchmark with a reproducible environment for evaluating task routing between AI agents.
US companies 10-K annual reports, split into sections and sentences, with sentiment labels.
Large cleaned corpus of real Russian-language reviews from 2023 for NLP tasks.
A set of paired English text continuations from humans and neural networks for style study and AI generation detector training.