HtmlDashboard
Models & Types · AiDotNet.TrainingMonitoring.Dashboard
Generates interactive HTML dashboards for training visualization.
Implements: ITrainingDashboard
How It Works
For Beginners: HtmlDashboard creates beautiful, interactive HTML reports that show your training progress. Unlike TensorBoard which requires running a server, these HTML files can be opened directly in any web browser.
Features:
- Interactive charts with zoom and pan (using Chart.js)
- Real-time loss and accuracy curves
- Confusion matrix heatmaps
- Histogram distributions
- ROC and PR curves
- Hyperparameter logging
Example usage:
Constructors
| Constructor | Summary |
|---|---|
HtmlDashboard(string logDirectory, string name) |
Creates a new HTML dashboard. |
Properties
| Property | Summary |
|---|---|
bool AutoSave { get; set; } |
Gets or sets whether to auto-save on each log operation. |
bool IsRunning { get; } |
|
int AutoSaveInterval { get; set; } |
Gets or sets the auto-save interval in number of logs. |
string LogDirectory { get; } |
|
string Name { get; } |
Methods
| Method | Summary |
|---|---|
Dictionary<string, List<HistogramDataPoint>> GetHistogramData() |
|
Dictionary<string, List<ScalarDataPoint>> GetScalarData() |
|
string GenerateReport(string outputPath) |
|
void Clear() |
|
void Dispose() |
Disposes the dashboard. |
void ExportTensorBoardFormat(string outputDirectory) |
|
void Flush() |
|
void LogConfusionMatrix(string name, long step, int[] matrix, string[] labels, DateTime? wallTime) |
|
void LogHistogram(string name, long step, double[] values, DateTime? wallTime) |
|
void LogHyperparameters(Dictionary<string, object> hyperparams, Dictionary<string, double> metrics) |
|
void LogImage(string name, long step, byte[] imageData, int width, int height, DateTime? wallTime) |
|
void LogModelGraph(string modelDescription) |
|
void LogPRCurve(string name, long step, double[] predictions, int[] labels, DateTime? wallTime) |
|
void LogROCCurve(string name, long step, double[] predictions, int[] labels, DateTime? wallTime) |
|
void LogScalar(string name, long step, double value, DateTime? wallTime) |
|
void LogScalars(Dictionary<string, double> scalars, long step, DateTime? wallTime) |
|
void LogText(string name, long step, string text, DateTime? wallTime) |
|
void Start() |
|
void Stop() |