{
  "_name_or_path": "schuler/experimental-JP47D18",
  "architectures": [
    "KPhi3ForCausalLM"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "do_layer_norm_before": true,
  "embd_pdrop": 0.0,
  "embed_size": 1280,
  "eos_token_id": 2,
  "ffn_dim": 1280,
  "hidden_act": "silu",
  "hidden_size": 1280,
  "initializer_range": 0.02,
  "intermediate_size": 5120,
  "max_position_embeddings": 4096,
  "min_channels_per_group": 256,
  "model_type": "kphi3",
  "num_attention_heads": 20,
  "num_hidden_layers": 4,
  "num_key_value_heads": 20,
  "original_max_position_embeddings": 4096,
  "pad_token_id": 0,
  "resid_pdrop": 0.0,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "sliding_window": null,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.46.3",
  "use_cache": true,
  "vocab_size": 32064
}