richard28039

About the 'lambda_ambient' meaning and high loss

1

Hello, Can anyone tell me the meaning of the 'lambda_ambient', when I was training the loss will increase slowly but not stop like below  thanks

Question about "gpt 4o" for manager worker, grouding model for local

1

> Hi @xyzhang626 , thank you for the support! > > We use `claude-3-7-sonnet-20250219` as our manager and worker, and use `bytedance-research/UI-TARS-72B-DPO` from HuggingFace as our grounding model. The hyperparams...