Constarium
← Search

Data · dataset · 2024

Developing Large Language Models for Quantum Chemistry Simulation Input Generation

Listed in IISH Dataverse

This repository encompasses all code used to run the experiments described in the study titled "Developing Large Language Models for Quantum Chemistry Simulation Input Generation".

Description

In addition to the code for our system architecture, we include the datasets described in the study, which can be used for further research. The repository also contains some generally helpful classes.

To reproduce the results from our study, refer to the Scripts folder, where we explain the scripts used to run our experiments and gather data. For more insight into the classes used and how to implement them in your own research, refer to the Classes folder. We for instance show how to easily use our rule-based system to generate different calculations.

Read the rest (2 more)

Additionally, you can inspect and extract the various datasets we used from the Data folder, where all available datasets are explained. The Orca Output folder stores all output files gathered from running ORCA calculations. One important note is that to use the code in this repository, you should configure your own OpenAI API key in your system path.

Moreover, to use RAG, one should scrape the ORCA input library with our provided script and add the ORCA manual to the Documents/Regular folder. We do not publish this here as we are not the writers.

Links

Where it is published

Catalogue records · 1

Topics

From keywords
Chemistry · Humanities · Physics
Provenance · 1 source records, 13 field assertions
SourceKeyLast seenRaw
IISH Dataversedoi:10.34894/WNRHA411 d agoJSON v1
FieldAssertionExtractorEvidence
concepts[field].anzsrc:group:3407enrichment · datasets iisg amsterdamtaxonomy-embedding@1.1.0title+keywords+description (75%)
concepts[field].dataverse_subject:chemistrysource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/subjects
concepts[field].dataverse_subject:computer-and-information-sciencesource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/subjects
concepts[field].local:field:chemistrymapping · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/subjects
concepts[field].local:field:humanitiesmapping · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/subjects
concepts[field].local:field:physicsmapping · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/subjects
concepts[method].local:method:simulationenrichment · datasets iisg amsterdamkeyword-concept-rules@1.0.0title+description (75%)
created_datesource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0
descriptionsource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/description
publication_datesource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0
titlesource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0/name
updated_datesource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0
version_labelsource · datasets iisg amsterdamconnector:datasets_iisg_amsterdam@1.0.0