⚙️Training Tasks

DeepQ AI Training is the place where the Neural Networks learns from training dataset (images & labels) . Users can train the Neural Network to do specific tasks and create their own AI application with personalized dataset.

In this section, users can create different training tasks, each designed for specific needs. These tasks include the settings, neural network selected and tuning method. It's a central hub that keeps track of and organizes all current and past training projects.

Task List

In the “Task” tab, you can view the training status of all tasks created.

• CREATION TIME: Filter your tasks by "CREATION TIME" or "FINISH TIME".

• STATUS: Filter your tasks by using the task status: “Running", "Waiting", "Finished", "Stopped", or "Error".

• SEARCH: Search your tasks using project name.

Task Detail

Clicking on any training task will take you to the task detail page, the information from top to bottom are:

• Task Training progress, including creation time, training duration & total GPU-hrs spent

• Training information: training dataset, model architecture and hyperparameters

• training report: real-time training accuracy & training curves (detailed pdf report downloadable)

• Dataset Statistics: Statistics of training data (training set & validation set)，in the form of both bar charts and table.

OPTION: Here task owners can edit task name, stop task (only when task is running/waiting) & delete task.