This model is a dummy model created for testing purposes only. It utilizes a custom configuration to explore various training scenarios and should not be used for production.
Configuration Highlights
The configuration of this dummy model is distinct from the original model in several key aspects:
Number of Layers: Reduced to 2, allowing for quicker tests of layer-specific behaviors.
Experts: Configured with 4 local experts and 2 experts per token, experimenting with the model's capacity to handle multiple expert inputs.
Hidden Size: Set at 512, this smaller size is suitable for testing the impact of network width.
Intermediate Size: Enlarged to 3579, to investigate how an increase in the size affects the model's ability to process information deeply.
