Melting Temperature Predictor Version 1
GNN model, accepts up to 4 elements.
Cite this model:
Qi-Jun Hong, A melting temperature database and a neural network model for melting temperature prediction, arXiv, 2021. Download.
Qi-Jun Hong, Sergey V Ushakov, Axel van de Walle, and Alexandra Navrotsky, Melting temperature prediction using a graph neural network model: From ancient minerals to new materials, PNAS, 2022. Download.
This model is currently deployed on Microsoft Azure and the Research Computing facilities at ASU. Due to network limit, computation may take up to 10 seconds.
Metrics
R2 training score: 0.99
R2 testing score: 0.96
Root mean square error, training: 75K
Root mean square error, testing: 138K
About this model
Batch Calculation
This model is also publicly accessible via command line, with a much shorter latency of ~1 second.
As shown in the following example, three materials (Ta, TaC, and TaHfC2) were computed in one API call. Users may include as many materials as they would like in one call.
Try this:
curl --request POST --url "http://206.207.50.58:5001/MT_ML_Qijun_Hong_Predict_curl" --header "content-type:application/json" --data '[{"elem1":"Ta","comp1":"1","elem2":"","comp2":"","elem3":"","comp3":"","elem4":"","comp4":""},{"elem1":"Ta","comp1":"1","elem2":"C","comp2":"1","elem3":"","comp3":"","elem4":"","comp4":""},{"elem1":"Ta","comp1":"1","elem2":"C","comp2":"2","elem3":"Hf","comp3":"1","elem4":"","comp4":""}]'
You will receive estimated melting temperatures in Kelvin from the model in seconds.
[{"elments ": ["Ta", "Vac", "Vac", "Vac"], "composition ": [100.0, 0.0, 0.0, 0.0], "melting temperature": 3588.9140625}, {"elments ": ["Ta", "C", "Vac", "Vac"], "composition ": [50.0, 50.0, 0.0, 0.0], "melting temperature": 4033.647216796875}, {"elments ": ["Ta", "C", "Hf", "Vac"], "composition ": [25.0, 50.0, 25.0, 0.0], "melting temperature": 4311.29541015625}]%