Skip to content

NeuralGraph is a powerful AI training data visualization tool designed to analyze and interpret training data for neural networks. It provides comprehensive visualizations of loss and accuracy trends over epochs, enabling users to gain deep insights into the training process and evaluate the performance of their AI models.

Notifications You must be signed in to change notification settings

Frost-Lord/NeuralGraph

Repository files navigation

NeuralGraph

NeuralGraph is an AI training data visualization tool that helps analyze and interpret training data for neural networks. It provides visualizations of loss and accuracy trends over epochs, allowing users to gain insights into the training process and assess the model's performance. When you call it simply create a popup app displaying the data.

NeuralGraph Preview

Features

  • Line charts for loss and accuracy trends over epochs
  • Calculation of average accuracy
  • Feedback message indicating the success or failure of AI training
  • Responsive and intuitive user interface
  • 3D Graph of you model
  • Nested Table to fully understand you model (Full Breakdown)

Coming soon:

  • Flow Charts
  • FAQ

Installation

Follow these steps to set up NeuralGraph:

  1. Install:
    npm i neuralgraph
  2. Call needed functions:
    const { GenerateGraph, updateGraph } = require("neuralgraph");
  3. Call this just before model.fit:
    GenerateGraph(model); //Parse the model directly into this to be able to see a 3D view of you model
  4. Update your model.fit callback:
    callbacks: {
      onEpochEnd: async (epoch, logs) => {
        console.log(
          `Epoch: ${epoch} Loss: ${logs.loss * 100} Accuracy: ${logs.acc}`
        );
       updateGraph(epoch, logs);
      };
    }

Data Format

NeuralGraph expects data in a specific format (default):

  • lossData: An array of loss values corresponding to each epoch.
  • accuracyData: An array of accuracy values corresponding to each epoch.

Example data format:

{
  "epoch": [0.5, 0.4, 0.3, 0.2, 0.1],
  "logs": [0.6, 0.7, 0.8, 0.9, 0.95]
}

Example Tensorflow.js code

const tf = require('@tensorflow/tfjs-node');
const { GenerateGraph, updateGraph } = require('neuralgraph');

const model = tf.sequential();
model.add(tf.layers.dense({ units: 1, inputShape: [1] }));
model.compile({ loss: 'meanSquaredError', optimizer: 'sgd', metrics: ['accuracy'] });

const xs = tf.tensor2d([-1, 0, 1, 2, 3, 4], [6, 1]);
const ys = tf.tensor2d([-3, -1, 1, 3, 5, 7], [6, 1]);

GenerateGraph(model);

model.fit(xs, ys, {
  epochs: 100,
  callbacks: {
    onEpochEnd: async (epoch, logs) => {
      console.log(`Epoch: ${epoch} Loss: ${logs.loss * 100} Accuracy: ${logs.acc}`);
      updateGraph(epoch, logs);
    }
  }
});

Contributing

Contributions are welcome! To contribute to NeuralGraph, follow these steps:

  1. Fork the repository.
  2. Create a new branch for your feature or bug fix.
  3. Make the necessary changes and commit them.
  4. Push your changes to your forked repository.
  5. Submit a pull request to the main repository.
  6. Please ensure that your code follows the existing coding style and conventions.

About

NeuralGraph is a powerful AI training data visualization tool designed to analyze and interpret training data for neural networks. It provides comprehensive visualizations of loss and accuracy trends over epochs, enabling users to gain deep insights into the training process and evaluate the performance of their AI models.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published