{
  "architectures": [
    "KosineForCausalLM"
  ],
  "attention_dropout": 0.0,
  "dtype": "float32",
  "hidden_dropout": 0.0,
  "hidden_size": 384,
  "initializer_range": 0.02,
  "intermediate_size": 1536,
  "max_position_embeddings": 512,
  "model_type": "fasty",
  "neftune_alpha": 0.0,
  "num_attention_heads": 8,
  "num_hidden_layers": 24,
  "num_key_value_heads": 2,
  "qk_norm": true,
  "rezero_init": 1.0,
  "rms_norm_eps": 1e-06,
  "rope_theta": 10000.0,
  "transformers_version": "5.0.0",
  "use_cache": false,
  "vocab_size": 32000
}