How to use Molecule Model Fine-Tuning Online

Fine-tune KERMT v2 for custom molecular property prediction.

Learn one or several molecular properties from measured SMILES datasets using NVIDIA KERMT v2. Choose regression or binary classification for all targets. Returns a validation-selected model, held-out predictions, metrics, and training history. At least 300 available compute credits are required. Jobs are limited to six hours; training time and credit use depend on dataset size and settings.

How Molecule Model Fine-Tuning Works

Prepare a CSV with ID, SMILES, and numeric target columns. List target names separated by commas. All targets use one task type: regression or binary classification (0/1). Blank measurements are masked. Each target needs at least two training labels and labels in validation and test; binary targets need both classes in every split. Keep target units consistent.

The default split allocates whole Bemis-Murcko scaffold groups toward 80% training, 10% validation, and 10% test. Actual fractions may differ. All acyclic molecules form one group. At least three scaffold groups are required. Alternatively fill Split on every row with train, validation, or test; scaffolds cannot cross splits.

Validation selects the checkpoint using mean training-normalized MAE for regression or mean log loss for classification. Test data is evaluated after selection. Review metrics.csv, training_history.csv, held_out_predictions.csv, and splits.csv. Use this job ID in Molecule Model Inference. The exported best_model.pt includes the learned weights and preprocessing; LICENSE.txt and NOTICE.txt accompany the trained model. Predictions require experimental validation for your application.

What is Neurosnap?

Neurosnap is the leading platform for bioinformatics and computational science focused on expanding access to powerful modeling and simulation tools. Because many state-of-the-art machine learning systems remain complex to install, configure, and scale, Neurosnap offers a clean, browser-based workspace that removes the burden of infrastructure management, dependency conflicts, and command-line tooling.

Built for biologists, chemists, and cross-disciplinary scientists, the platform enables advanced computational workflows without requiring expertise in software engineering or cloud architecture. Researchers can launch analyses through an intuitive interface, connect programmatically through a comprehensive API, and rely on automated resource management to scale workloads efficiently. By taking care of the underlying compute and operational complexity, Neurosnap allows teams to devote their energy to scientific progress and faster iteration. Security and data protection remain foundational principles, with clear safeguards outlined in our Terms of Use and Privacy Policy to ensure your work stays protected.

Advancing Discovery with Molecule Model Fine-Tuning on Neurosnap

Using Molecule Model Fine-Tuning on Neurosnap could drastically accelerate .

How to Use Molecule Model Fine-Tuning on Neurosnap

To harness the capabilities of Molecule Model Fine-Tuning, researchers can follow this streamlined workflow within Neurosnap:

Access Neurosnap: Start by logging in to the Neurosnap website.
Select Tool: From the list of available tools, choose Molecule Model Fine-Tuning.
Provide Inputs: Provide all the inputs specified within the submission panel and optionally configure the tool as desired.
Run Tool: Submit the Molecule Model Fine-Tuning job and Neurosnap will execute it in the cloud, automatically notifying you as soon as your results are ready.
Review Output: Explore your results through rich visualizations, including figures, plots, and interactive views designed to help you analyze findings with clarity and confidence.

Citations

Please cite the original work when using Molecule Model Fine-Tuning in publications or research outputs.

Multitask finetuning and acceleration of chemical pretrained models for small molecule drug property prediction. https://arxiv.org/abs/2510.12719

Contrastive KERMT. https://arxiv.org/abs/2606.11508

Neurosnap Inc. (2022). Neurosnap: An online platform for computational biology and chemistry. Available at: https://neurosnap.ai/

Similar Services

Explore related tools that support similar research workflows:

Molecule Model Inference

Admetica

ChemBounce

Permeability and Efflux Prediction

Proudly supporting 50,000+ scientists worldwide, including 7,000+ leading biotech and global biopharma organizations.