HF RL Explorer

Task: Demonstrate the Difference Between PyTorch Adam with weight decay and AdamW

Task: Demonstrate the Difference Between PyTorch Adam with weight decay and AdamW: a task in Terminal-Lego-15k (Harbor dataset). Write a Python script that empirically demonstrates the difference between torch.optim.Adam(weight decay=0.01) and torch.optim.AdamW(weight decay=0.01) by showing how…

The task

Write a Python script that empirically demonstrates the difference between `torch.optim.Adam(weight_decay=0.01)` and `torch.optim.AdamW(weight_decay=0.01)` by showing how each optimizer updates model parameters differently.

Part of PrimeIntellect/Terminal-Lego-15k.