Search results

The confused the concepts of fixed decay and decoupled decay

Status: Open.

#39 In lucidrains/lion-pytorch;

· zzp1012 opened on Oct 24, 2025

Using Triton with PyTorch 2.0 for AMP training results in tensors containing inf values.

Status: Open.

#35 In lucidrains/lion-pytorch;

· DrRyanHuang opened on Feb 17, 2024

Always getting NaNs in long training

Status: Open.

#33 In lucidrains/lion-pytorch;

· danbochman opened on Dec 5, 2023

AMD ROCM versions

Status: Open.

#32 In lucidrains/lion-pytorch;

· bennmann opened on Aug 6, 2023

Convergence guarantees for Lion

Status: Open.

#30 In lucidrains/lion-pytorch;

· xnchxy opened on Jul 20, 2023

Do you have the actual weights trained from the paper?

Status: Open.

#29 In lucidrains/lion-pytorch;

· tr1-ai opened on Jun 25, 2023

Same amount of VRAM is taken as in AdamW

Status: Open.

#22 In lucidrains/lion-pytorch;

· VCasecnikovs opened on Apr 3, 2023

add an 8-bit version with bitsandbytes

enhancementNew feature or requestNew feature or request

Status: Open.

#17 In lucidrains/lion-pytorch;

· lucidrains opened on Mar 8, 2023

Learning rate scaling for distributed training?

Status: Open.

#8 In lucidrains/lion-pytorch;

· RahulBhalley opened on Feb 18, 2023