---
pretty_name: Gradients Math Comparison
task_categories:
  - question-answering
  - text-generation
tags:
  - conversational
  - math
  - jsonl
configs:
  - config_name: default
    data_files:
      - split: train
        path: math_train.jsonl
      - split: test
        path: math_test.jsonl
---

# Gradients math comparison

This public copy contains 7,913 training rows and 879 test rows. Each row has
exactly one user turn followed by one assistant turn under `conversations`. The local
source uses the equivalent top-level key `messages`; only that key was renamed for
compatibility with Gradients baseline preparation.