README.md

GitHub Issues Dataset Card

Author / Maintainer: @xanderIV
Point of Contact: @xanderIV

---

Dataset Description

Dataset Summary

GitHub Issues is a dataset consisting of GitHub issues and pull requests associated with the 🤗 Datasets repository:
https://github.com/huggingface/datasets

This dataset is curated and documented by @xanderIV for educational and research purposes.

It can be used for:

• Semantic search
• Multilabel text classification
• Automated issue triaging
• Topic modeling
• LLM fine-tuning
• Retrieval-Augmented Generation (RAG) experiments

The dataset contains English-language technical discussions related to NLP, computer vision, speech, multimodal ML systems, and ML infrastructure.

This dataset is particularly relevant for:

• Developer assistant systems
• LLM-powered support automation
• DevOps / MLOps / LLMOps workflows
• Research in applied ML systems

---

Supported Tasks and Leaderboards

1. Text Classification (text-classification)

The dataset can be used for multilabel text classification, where a model predicts one or more labels (e.g., bug, enhancement, documentation) for each issue.

Typical metrics:

• F1 score
• Accuracy
• Precision
• Recall

Suggested models:

• distilbert-base-uncased
• roberta-base
• microsoft/deberta-v3-base

---

2. Information Retrieval (information-retrieval)

The dataset can be used for semantic search, where the task is to retrieve the most relevant GitHub issue given a user query.

Typical metrics:

• MRR (Mean Reciprocal Rank)
• Recall@k

Suggested models:

• sentence-transformers/all-MiniLM-L6-v2
• BAAI/bge-base-en-v1.5
• intfloat/e5-base

---

3. Issue Triaging (other:issue-triaging)

The dataset can be used for automated issue routing and classification.

The task consists of:

• Predicting labels
• Suggesting maintainers
• Routing issues to appropriate teams

Metrics:

• Classification accuracy
• Routing precision

---

4. LLM Fine-Tuning (other:llm-finetuning)

The dataset can be used to fine-tune large language models for:

• Developer assistants
• Issue summarization
• Pull request review generation
• Support automation

Suggested models:

• meta-llama/Llama-3-8B-Instruct
• mistralai/Mistral-7B-Instruct-v0.2
• Qwen/Qwen2.5-7B-Instruct

---

Languages

• Primary language: English (en, BCP-47)

Text characteristics:

• Technical discussions
• Developer communication
• Bug reports
• Code snippets (Python, YAML, JSON, etc.)
• Configuration files

Language style: semi-formal, domain-specific (software engineering / ML infrastructure).

---

Dataset Structure

Data Instances

Example instance:

{
  "issue_id": 12345,
  "title": "Dataset loading fails with streaming=True",
  "body": "When trying to load the dataset with streaming enabled...",
  "labels": ["bug", "datasets"],
  "author": "username",
  "created_at": "2023-06-10T14:32:00Z",
  "comments": [
    {
      "author": "maintainer",
      "text": "Can you provide the stack trace?"
    }
  ]
}