validation loss
Validation loss is a metric used to evaluate how well a machine learning model is performing on unseen data during the training process. It measures the error of the model on a validation dataset that is separate from the training data, helping to assess the model's generalization capability.
- Coarse-to-fine Q-Network with Action Sequence for Data-Efficient Reinforcement Learning
- Diffusion Beats Autoregressive in Data-Constrained Settings
- LayerIF: Estimating Layer Quality for Large Language Models using Influence Functions
- Scale-invariant attention
- Scaling Law with Learning Rate Annealing
- Tensor-Parallelism with Partially Synchronized Activations