Qwen3-4B-Instruct-2507 fully fine-tuned on the panasonic_qa_claude_v1 config of Industrial-Instruction — the split generated with Claude-Opus-4.6 using an identical pipeline, prompts and scenario definitions.

A companion model trained on the open-weight-generated split is available at Parssky/industrial-instruction-qwen4b.

Paper

Industrial-Instruction: An End-to-End Framework for Building Instruction-Tuning and Benchmark Datasets from Industrial Technical Reports

Base Model

Qwen3-4B-Instruct-2507

Results

Panasonic benchmark (panasonic_qa_claude_v1 test split, 1,000 items):

Configuration	F1	Jaccard	Set-Match Acc.
Base, with RAG	58.55%	54.00%	40.90%
Fine-tuned, with RAG	72.66%	68.88%	56.40%
Base, no RAG	59.24%	54.71%	41.60%
Fine-tuned, no RAG	72.72%	68.93%	56.40%

General knowledge (MMLU): 72.13% base → 72.08% fine-tuned (−0.05) — essentially no forgetting. Several subjects improve: global_facts (+8.00), college_mathematics (+7.00), machine_learning (+5.36), security_studies (+4.49), international_law (+3.30).

FailureSensorIQ: AccOrgIBM 34.0% → 49.6%, F1-Macro 40.0% → 33.5%, F1-Micro 66.0% → 50.3%.

Note: the two models are evaluated on different held-out test splits, each generated by its own generator model. The baseline scores therefore differ (28.5% vs. 40.9%), and the two improvement magnitudes are indicative rather than a controlled head-to-head comparison.

Limitations

Accuracy on perturbed (rephrased) FailureSensorIQ questions is 0%, both before and after fine-tuning. The training data contains no paraphrased question variants, so this model should not be relied on where question phrasing varies.
Fine-tuning costs 1.26 points of MMLU accuracy, concentrated in Humanities (moral-reasoning subjects drop ~10 points).
Source documentation comes from a single manufacturer; terminology may not transfer.

Intended Use

Research and benchmarking for industrial retrieval-augmented generation (RAG), evidence integration, and technical-domain QA.

Notes

Inference-ready weights and tokenizer files only; training checkpoints and optimizer states are intentionally excluded.

Source Code

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

Parssky/industrial-instruction-qwen4b-claude

Industrial-Instruction Qwen3-4B (Claude-generated data)