davidx08's picture
|
download
raw
1.14 kB
metadata
language:
  - en
license: apache-2.0
task_categories:
  - text-generation
  - question-answering
tags:
  - code
  - instruction-tuning
  - sharegpt
  - fine-tuning
  - python
  - javascript
  - go
size_categories:
  - 100K<n<1M

Coding Master Dataset

Overview

A large-scale coding instruction-tuning dataset in ShareGPT conversational format, assembled from multiple open sources and deduplicated.

Records: 766,987 Format: JSONL / ShareGPT License: Apache 2.0

Sources

  • CodeX-2M-Thinking (430,542 records)
  • python-code-dataset-500k (559,515 records)
  • StackPulse high-quality subset (20,205 records)
  • CodeFeedback-Filtered-Instruction (156,525 records)
  • secure_programming_dpo (4,656 records)
  • Go coding JSONL files (267,461 records)

Schema

Field Type Description
conversations list ShareGPT format human/gpt turns
source string Origin dataset
source_id string Unique record identifier
metadata_json string Additional metadata

Usage

from datasets import load_dataset
ds = load_dataset("Voidreaper2026/coding-master-dataset")

License

Apache 2.0

Xet Storage Details

Size:
1.14 kB
·
Xet hash:
7b3ee7c2e81d686458fadd6f14941ae189801b6fec11a274978a9fd7befb82a3

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.