original/config.json

{
  "num_hidden_layers": 6,
  "num_experts": 8,
  "experts_per_token": 4,
  "vocab_size": 1024,
  "hidden_size": 128,
  "intermediate_size": 128,
  "swiglu_limit": 7.0,
  "head_dim": 32,
  "num_attention_heads": 4,
  "num_key_value_heads": 1,
  "sliding_window": 128,
  "initial_context_length": 1024,
  "rope_theta": 150000.0,
  "rope_scaling_factor": 4.0,
  "rope_ntk_alpha": 1.0,
  "rope_ntk_beta": 32.0
}