README.md

metadata

license: apache-2.0
tags:
  - chemistry
  - precite
  - chemberta
datasets:
  - blainetrain/precite-dataset-FLP-Reactivity-Predictor-1-1
base_model: seyonec/ChemBERTa-zinc-base-v1
model-index:
  - name: FLP-Reactivity-Predictor-1-1
    results:
      - task:
          type: molecular-property-prediction
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.25
          - name: F1
            type: f1
            value: 0.25
          - name: Precision
            type: precision
            value: 0.25
          - name: Recall
            type: recall
            value: 0.25

FLP Reactivity Predictor 1.1

A chemistry prediction model fine-tuned on Precite platform.

Model Details

• Base Model: seyonec/ChemBERTa-zinc-base-v1
• Fine-tuned On: 14 training samples, 4 validation samples (80/20 split)
• Task: Molecular property prediction (3 classes)
• Epochs: 3
• Training Date: 2026-02-06

Performance Metrics (20% Holdout Test Set)

Metric	Value
Accuracy	0.2500
F1 Score	0.2500
Precision	0.2500
Recall	0.2500
Training Loss	1.3063

Label Classes

• no_reaction
• reaction_occurred
• unknown

Usage

This model can be queried through the Precite platform for FLP chemistry predictions.

from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained("blainetrain/FLP-Reactivity-Predictor-1-1")
tokenizer = AutoTokenizer.from_pretrained("blainetrain/FLP-Reactivity-Predictor-1-1")

Training Data

See the associated dataset: blainetrain/precite-dataset-FLP-Reactivity-Predictor-1-1