what is the value you use lambda_vel and lambda_fc?

[sinhasaptarshi]

For training the text to motion model, what values do you use for lambda_vel and lambda_fc (ie the lambda for foot contact loss and joint velocity loss)? or dont you use it for text to motion?

[GuyTevet]

Those are not used in text-to-motion since the velocities and contacts are already embedded into the data representation.