SeqReg pretrained models

SeqReg is an open sequence-regression software package for experimental pool-boiling heat-flux prediction using hydrophone, acoustic-emission (AE) hit, and optical-image inputs. The canonical software source, installation instructions, tutorials, and implementation history are maintained at cldunlap73/SeqReg.

This repository provides the pretrained artifacts released with SeqReg. It is not a hosted inference endpoint.

Included artifacts

File Intended input and role SHA-256
HydReg.joblib Hydrophone sequence regression; the canonical documentation specifies FFT preprocessing and SeqLen=4000. FBEB30D598EC7BAB58F0DD981BDB3A7C2149FC2EACC04E0E74DF60DAAFA4DAF8
Hit2Flux_weights.h5 AE-hit sequence regression; the canonical documentation specifies FFT preprocessing, SeqLen=25, and sequence output. 9942A030D9399D2D6F4AFA9A163B95B0112855EF78CE5FC99BD3975B9629A86D
ImgReg.hdf5 Optical-image sequence-regression artifact. 198254381899CEC72BA7366ECC4BEE3AD1C67D5E32D15F281DF23D90287C747F
ImagePCA.pkl PCA artifact required by ImgReg; canonical configuration specifies PCAnpy, pcskeep=40, and SeqLen=200. B565E08A80B8F85ED834D485AC295FBCE5876EE5BC1C46269997A29112DB4C0D

Data and reproducibility

The self-contained model-and-test-data package is available at UARK-NED3/BoilingBench-SeqReg. The original public SeqReg OSF project remains the release record referenced by the software documentation.

Use the original SeqReg README for required dependencies, data layout, loading logic, and tutorials. The Hugging Face Dataset retains the supplied test-data directory structure; do not infer a model-input convention from filenames alone.

Validation status and limitations

The four local artifacts were SHA-256 verified before upload. They were not loaded or re-evaluated in this release workflow. These models were developed for the documented pool-boiling measurement modalities and configurations; evaluate transfer to another facility, fluid, sensor chain, surface, operating regime, or transient protocol before research use.

joblib and pickle artifacts can execute code when deserialized. Load only files obtained from this verified repository and in a controlled environment.

Attribution and licensing

SeqReg and the linked OSF project are released under Apache-2.0. Cite the applicable publication and use the canonical repository for software issues and updates.

  • C. Dunlap, H. Pandey, E. Weems, and H. Hu, Nonintrusive Heat Flux Quantification Using Acoustic Emissions During Pool Boiling, Applied Thermal Engineering, 2023. https://doi.org/10.1016/j.applthermaleng.2023.120558
  • C. Dunlap, C. Li, H. Pandey, and H. Hu, Hit2Flux: A Machine Learning Framework for Boiling Heat Flux Prediction Using Hit-Based Acoustic Emission Sensing, AI in Thermal Fluids, 2025. https://doi.org/10.1016/j.aitf.2025.100002
  • C. Dunlap, C. Li, H. Pandey, Y. Sun, and H. Hu, A Temporal-Spatial Framework for Efficient Heat Flux Monitoring of Transient Boiling, IEEE Transactions on Instrumentation and Measurement, 2024. https://doi.org/10.1109/TIM.2024.3460944
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support