A Deep Learning Dataloader with Shared Data Preparation

Keyphrases

• Data Sharing 100%
• Data Preparation 100%
• Training Intensity 100%
• Deep Learning 100%
• Sampling Algorithm 66%
• High Efficiency 33%
• Loading Method 33%
• Redundancy 33%
• Low Sampling Rate 33%
• Cache Replacement 33%
• Parallel Jobs 33%
• Strong Constraint 33%
• Speed Improvement 33%
• Random Data 33%
• Tree Data Structures 33%
• Deep Learning Model 33%
• Cache Thrashing 33%
• ResNet18 33%
• Data Shuffling 33%
• Data Loading 33%
• Flexible Constraints 33%
• Centralized Cache 33%

Computer Science

• Deep Learning 100%
• Data Preparation 100%
• Sampling Algorithm 66%
• Data Structure 33%
• Input/Output 33%
• Common Practice 33%
• Deep Neural Network 33%
• Deep Learning Model 33%
• Data Loading 33%