tasksource/bigbench
Multiple ChoiceQuestion AnsweringText ClassificationText GenerationZero Shot ClassificationENBenchmark
The tasksource/bigbench dataset is a EN multiple choice resource from tasksource at 2023.
📊 This dataset is used as an LLM benchmark. See model leaderboards →
About tasksource/bigbench
BIG-Bench but it doesn't require the hellish dependencies (tensorflow, pypi-bigbench, protobuf) of the official version.
dataset = load_dataset("tasksource/bigbench",'movie_recommendation')
Code to reproduce:
https://colab.research.google.com/dri...
Details
- Task
- Multiple Choice, Question Answering, Text Classification, Text Generation, Zero Shot Classification
- Language
- EN
- Format
- Parquet
- Rows / instances
- N/A
- Creator
- tasksource
- Year
- 2023