JournalFit NLP Data Repository Package

The repository contains processed and analyzed data files generated during the JournalFit-NLP computational evaluation. CSV files provide tabular summaries of dataset composition, retrieval ablation, reranker blending, latency, quantization fidelity, resource monitoring, hardware/software environment, and expanded-pool evaluation metrics. JSON files contain machine-readable metric outputs for baseline, fine-tuned, expanded-pool, and reranking experiments. YAML files document model training and evaluation configurations. PNG and PDF figures visualize the architecture, development flow, retrieval ablations, reranker blending, reranker failure diagnosis, training/thermal monitoring, quantization trade-off, and CPU latency profiles. No private manuscripts, identifiable user records, social media data, or animal data are included.