---
license: apache-2.0
task_categories:
- text-generation
- question-answering
language:
- en
tags:
- math
- code
- bench
- benchmark
- banana
- bananamind
- knowledge
- arithmetic
- SLM
- instruct
- instruction
- hard
- easy
- mathematics
- programming
size_categories:
- 1K<n<10K
---

# ArrowBench-Instruct

**ArrowBench-Instruct** is a big expansion to the bananamind instruct 1.1 benchmark, featuring a 3.3 times bigger benchmark, harder questions, and more diverse questions.

Use the python script inside this dataset repo to test **your** instruct model's results on this benchmark!

## Dataset Details

- **Curated by:** AInion
- **Funded by:** ArrowLabs
- **Shared by:** MoreThought
- **License:** apache 2.0

## Dataset Source

**BananaMind/BananaMind-Instruct-Bench-1.1**

## Extra

For more information about the script, data format, and categories, you can go to the original **BananaMind/BananaMind-Instruct-Bench-1.1**.