Powered by OpenAIRE graph
Found an issue? Give us feedback
ZENODOarrow_drop_down
ZENODO
Software . 2026
License: CC BY
Data sources: Datacite
ZENODO
Software . 2026
License: CC BY
Data sources: Datacite
ZENODO
Software . 2026
License: CC BY
Data sources: Datacite
ZENODO
Software . 2026
License: CC BY
Data sources: Datacite
versions View all 4 versions
addClaim

BERTology of Molecular Property Prediction (Data Size Effect: Small-BERT, Part 1)

Authors: Mostafanejad, Mohammad; Saxe, Paul; Crawford, T. Daniel;

BERTology of Molecular Property Prediction (Data Size Effect: Small-BERT, Part 1)

Abstract

Data Size Effects on Pre-Training Small-BERT (Part 1) This record pertains to Data Size Effects on Pre-Training Experiments. It includes the following files: small-binidx-0-ms-1234-ds-1234.tar.gz small-binidx-0-ms-1234-ds-2345.tar.gz small-binidx-0-ms-2345-ds-1234.tar.gz small-binidx-1-ms-1234-ds-1234.tar.gz small-binidx-1-ms-1234-ds-2345.tar.gz small-binidx-1-ms-2345-ds-1234.tar.gz small-binidx-2-ms-1234-ds-1234.tar.gz small-binidx-2-ms-1234-ds-2345.tar.gz small-binidx-2-ms-2345-ds-1234.tar.gz small-binidx-3-ms-1234-ds-1234.tar.gz small-binidx-3-ms-1234-ds-2345.tar.gz small-binidx-3-ms-2345-ds-1234.tar.gz Each tar file contains all model artifacts (checkpoints, random-number generator states, optimizer states etc.), training logs (Tensorboard, MLFlow and Weights & Biases), and evaluation results, configuration files, run scripts, SLURM sbatch driver scripts, and any additional artifacts generated during the experiments. Preprint: https://doi.org/10.48550/arXiv.2603.13627

Related Organizations
Keywords

Masked language modeling, Molecular property prediction, ADME, Maksed language modeling, Pretraining, Finetuning, Chemical language models, BERT

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average