---
pretty_name: Industrial-Instruction Dataset
language:
- en
license: other
task_categories:
- question-answering
- text-generation
tags:
- industrial
- rag
- benchmark
- multiple-choice
configs:
- config_name: panasonic_qa_v1
  data_files:
  - split: train
    path: panasonic_qa_v1/train-*
  - split: test
    path: panasonic_qa_v1/test-*
- config_name: panasonic_qa_claude_v1
  data_files:
  - split: train
    path: panasonic_qa_claude_v1/train-*
  - split: test
    path: panasonic_qa_claude_v1/test-*
- config_name: corpus_panasonic_md_v0_1
  data_files:
  - split: train
    path: corpus_panasonic_md_v0_1/train-*
- config_name: panasonic_v0_0
  data_files:
  - split: train
    path: panasonic_v0_0/train-*
dataset_info:
- config_name: panasonic_qa_claude_v1
  features:
  - name: question
    dtype: string
  - name: answer
    list: string
  - name: documents
    list: string
  splits:
  - name: train
    num_bytes: 168144704
    num_examples: 25252
  - name: test
    num_bytes: 6607933
    num_examples: 1000
  download_size: 162307467
  dataset_size: 174752637
- config_name: panasonic_v0_0
  features:
  - name: page_no
    dtype: int64
  - name: input_height
    dtype: int64
  - name: input_width
    dtype: int64
  - name: layout_info_path
    dtype: string
  - name: layout_image_path
    dtype: string
  - name: md_content_path
    dtype: string
  - name: filtered
    dtype: bool
  - name: file_path
    dtype: string
  - name: md_content
    dtype: string
  splits:
  - name: train
    num_bytes: 993352721
    num_examples: 7525
  download_size: 986131439
  dataset_size: 993352721
---

# Industrial-Instruction Dataset

Industrial-Instruction provides benchmark and training-ready QA instances derived from industrial technical reports, designed to evaluate robustness under realistic retrieval conditions. Samples are grounded in retrieved evidence and include irrelevant retrieval, single-/multi-document support, and single-/multi-document answer settings.

## Paper

[Industrial-Instruction: An End-to-End Framework for Building Instruction-Tuning and Benchmark Datasets from Industrial Technical Reports](https://huggingface.co/papers/2608.22817)

arXiv: https://arxiv.org/abs/2608.22817

## Configs

| Config | Records | Description |
|---|---|---|
| `panasonic_qa_v1` | 12,557 train / 1,000 test | QA data generated with the open-weight Qwen3-30B-A3B-Instruct model. |
| `panasonic_qa_claude_v1` | 25,252 train / 1,000 test | QA data generated with Claude-Opus-4.6, same pipeline and prompts. |
| `corpus_panasonic_md_v0_1` | — | Retrieval corpus: layout-preserved Markdown extracted from the source PDFs. |
| `panasonic_v0_0` | 7,525 | Raw, unfiltered page-level extractions before quality filtering. |

Each QA record is grounded in five query–document scenarios (r0–r4): irrelevant retrieval, single-/multi-document support, and single-/multi-document answer.

## Usage

```python
from datasets import load_dataset

# QA data generated with the open-weight Qwen3-30B-A3B-Instruct model
qa = load_dataset("Parssky/industrial-instruction-dataset", "panasonic_qa_v1")
# DatasetDict: train (12,557) / test (1,000)

# QA data generated with Claude-Opus-4.6
qa_claude = load_dataset("Parssky/industrial-instruction-dataset", "panasonic_qa_claude_v1")
# DatasetDict: train (25,252) / test (1,000)

# Retrieval corpus (layout-preserved Markdown pages)
corpus = load_dataset("Parssky/industrial-instruction-dataset", "corpus_panasonic_md_v0_1")

# Raw, unfiltered page-level extractions
raw = load_dataset("Parssky/industrial-instruction-dataset", "panasonic_v0_0")
```

### Record format

Each QA record has three fields:

| Field | Type | Description |
|---|---|---|
| `question` | `string` | The question, with its five answer options (A–E) inline. |
| `answer` | `list[string]` | Correct option letter(s), e.g. `["A"]` or `["B", "C"]`. Some questions have multiple correct answers. |
| `documents` | `list[string]` | Source passages retrieved from the Panasonic corpus when the item was generated. |

```python
ex = qa["test"][0]
print(ex["question"])    # "... which performance characteristic should be prioritized ...
                         #  A Thermal shock resistance  B Solderability  ..."
print(ex["answer"])      # ["A"]
print(ex["documents"])   # ["Current Sensing Resistors, Metal Plate Type\n\n## Performance ..."]
```

### Evaluation

Answers are sets, not ordered strings, so exact-match scoring is misleading
(`["A","B"]` vs `["B","A"]`). The paper scores with Set-Match Accuracy, F1 and
Jaccard similarity. Benchmark scripts:
[`package_benchmark_panasonic`](https://github.com/parssky/industrial-instruction/tree/main/package_benchmark_panasonic).

## Models trained on this dataset

- [Parssky/industrial-instruction-qwen4b](https://huggingface.co/Parssky/industrial-instruction-qwen4b) — trained on `panasonic_qa_v1`
- [Parssky/industrial-instruction-qwen4b-claude](https://huggingface.co/Parssky/industrial-instruction-qwen4b-claude) — trained on `panasonic_qa_claude_v1`


## Intended Use

This dataset is for research on industrial retrieval-augmented generation (RAG), evidence integration, and instruction tuning for technical-domain QA.

## Notes

- Derived from publicly available industrial technical documentation published by Panasonic Corporation.
- Use should follow source-document terms and applicable data-use restrictions.

## Source Code

GitHub repository: https://github.com/parssky/industrial-instruction

## Citation

```bibtex
@misc{parsa_bakhtiari_2026,
	author       = { Parsa Bakhtiari and Hassan Bashiri and Alireza Khalilipour and Masoud Nasiripour and Moharram Challenger },
	title        = { industrial-instruction-dataset (Revision 7eadea0) },
	year         = 2026,
	url          = { https://huggingface.co/datasets/Parssky/industrial-instruction-dataset },
	doi          = { 10.57967/hf/10098 },
	publisher    = { Hugging Face }
}
```