Sovereign 8s.is mirror of standardgalactic/PADP
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
2026-01-26 23:51:35 +08:00
convs first commit 2026-01-26 23:13:14 +08:00
base.py first commit 2026-01-26 23:13:14 +08:00
data.py first commit 2026-01-26 23:13:14 +08:00
data_manager.py first commit 2026-01-26 23:13:14 +08:00
DynamicDataPruning.py first commit 2026-01-26 23:13:14 +08:00
factory.py first commit 2026-01-26 23:13:14 +08:00
icarl.json first commit 2026-01-26 23:13:14 +08:00
icarl.py first commit 2026-01-26 23:13:14 +08:00
inc_net.py first commit 2026-01-26 23:13:14 +08:00
main.py first commit 2026-01-26 23:13:14 +08:00
README.md Update README.md 2026-01-26 23:51:08 +08:00
requirements.txt first commit 2026-01-26 23:22:20 +08:00
toolkit.py first commit 2026-01-26 23:13:14 +08:00
trainer.py first commit 2026-01-26 23:13:14 +08:00

PADP: Progressive and Adaptive Data Pruning for Efficient Incremental Learning

PADP is a progressive and adaptive data pruning method for incremental learning, which dynamically evaluates sample difficulty and its difficulty changes during training to enable adaptive sample selection for each incremental learning task. Specifically, we propose two metrics, the instant difficulty score and the difficulty variation score. The former evaluates the learning difficulty of a sample, while the latter evaluates the variation in difficulty over a training interval. These two metrics are combined to guide pruning decisions. To prevent certain classes from being completely removed, we also introduce a class-balance retention mechanism.

🚀 Quick start

  1. Clone the repository:
git clone https://github.com/duanbiqing/PADP.git
cd PADP
  1. Install dependencies:
pip install -r requirements.txt
  1. Run the experiment:
python main.py