Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
sentiment-classification
Languages:
English
Size:
10K - 100K
License:
| { | |
| "default": { | |
| "description": "The Stanford Sentiment Treebank consists of sentences from movie reviews and\nhuman annotations of their sentiment. The task is to predict the sentiment of a\ngiven sentence. We use the two-way (positive/negative) class split, and use only\nsentence-level labels.\n", | |
| "citation": "@inproceedings{socher2013recursive,\n title={Recursive deep models for semantic compositionality over a sentiment treebank},\n author={Socher, Richard and Perelygin, Alex and Wu, Jean and Chuang, Jason and Manning, Christopher D and Ng, Andrew and Potts, Christopher},\n booktitle={Proceedings of the 2013 conference on empirical methods in natural language processing},\n pages={1631--1642},\n year={2013}\n}\n", | |
| "homepage": "https://nlp.stanford.edu/sentiment/", | |
| "license": "Unknown", | |
| "features": { | |
| "idx": { | |
| "dtype": "int32", | |
| "_type": "Value" | |
| }, | |
| "sentence": { | |
| "dtype": "string", | |
| "_type": "Value" | |
| }, | |
| "label": { | |
| "names": [ | |
| "negative", | |
| "positive" | |
| ], | |
| "_type": "ClassLabel" | |
| } | |
| }, | |
| "builder_name": "sst2", | |
| "dataset_name": "sst2", | |
| "config_name": "default", | |
| "version": { | |
| "version_str": "2.0.0", | |
| "major": 2, | |
| "minor": 0, | |
| "patch": 0 | |
| }, | |
| "splits": { | |
| "train": { | |
| "name": "train", | |
| "num_bytes": 4681603, | |
| "num_examples": 67349, | |
| "dataset_name": null | |
| }, | |
| "validation": { | |
| "name": "validation", | |
| "num_bytes": 106252, | |
| "num_examples": 872, | |
| "dataset_name": null | |
| }, | |
| "test": { | |
| "name": "test", | |
| "num_bytes": 216640, | |
| "num_examples": 1821, | |
| "dataset_name": null | |
| } | |
| }, | |
| "download_size": 3331058, | |
| "dataset_size": 5004495, | |
| "size_in_bytes": 8335553 | |
| } | |
| } |