hugg1ngfac3/deepseek-r1-tiny

This is a slice of deepseek-ai/DeepSeek-R1 for testing purpose:

1. n_routed_experts is updated from 256 to 16, i.e. there're 16 experts per layer
2. only layers.2 (dense MLP) and layers.3 (MoE) are kept

This way the model size is minimized, but people can still use this checkpoint for (model loader, MoE forward path) testing