keurcien
Had the same issue today, tried to run finetune/adapter_v2 for Falcon 7B on a NVIDIA L4 (24 GB RAM) ``` Loading model 'checkpoints/tiiuae/falcon-7b/lit_model.pth' with {'org': 'tiiuae', 'name': 'falcon-7b', 'block_size': 2048,...
Should the values of token expired_at or created_at be persisted in the token storage layer? Can we make this explicit in the token storage protocol? Currently users need to guess...
Agreed, but this fix should be enough? Combined with update on expiry logic using expired_at. ```python async def _initialize(self): """Initialize and properly set token expiry from stored tokens.""" self.context.current_tokens =...