config.json

{
  "architectures": [
    "TTTForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "modeling_ttt.TTTConfig",
    "AutoModel": "modeling_ttt.TTTModel",
    "AutoModelForCausalLM": "modeling_ttt.TTTForCausalLM"
  },
  "bos_token_id": 1,
  "conv_kernel": 4,
  "dtype": "bfloat16",
  "eos_token_id": 2,
  "hidden_act": "silu",
  "hidden_size": 768,
  "initializer_range": 0.02,
  "intermediate_size": 2048,
  "max_position_embeddings": 2048,
  "mini_batch_size": 16,
  "model_type": "ttt",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pre_conv": true,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-06,
  "rope_theta": 10000.0,
  "scan_checkpoint_group_size": 0,
  "share_qk": true,
  "transformers_version": "4.57.6",
  "ttt_base_lr": 1.0,
  "ttt_layer_type": "linear",
  "use_cache": true,
  "use_gate": true,
  "vocab_size": 32000
}