step_000000012000/config.json

{
  "architectures": [
    "ZebraModel"
  ],
  "attention_bias": false,
  "attention_dropout": false,
  "bos_token_id": 0,
  "dtype": "float32",
  "eos_token_id": 0,
  "head_dim": 64,
  "head_intermediate_size": 2560,
  "hidden_act": "silu",
  "hidden_size": 1024,
  "initializer_range": 0.02,
  "intermediate_size": 2560,
  "latent_size": 128,
  "loss_ema_beta": 0.75,
  "max_position_embeddings": 8192,
  "mlp_bias": false,
  "model_type": "zebra",
  "mu_scale": 2.0,
  "num_attention_heads": 16,
  "num_hidden_layers": 24,
  "num_key_value_heads": 4,
  "num_mixture_components": 8,
  "pad_token_id": 0,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 100000,
  "tie_word_embeddings": false,
  "transformers_version": "4.57.1",
  "use_cache": false,
  "vocab_size": 1,
  "whiten_embs": true,
  "z_ar_steps": 16,
  "z_length": 384
}