config.json

{
  "architectures": [
    "GemmaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 2,
  "dtype": "float32",
  "eos_token_id": 1,
  "head_dim": 64,
  "hidden_act": "gelu_pytorch_tanh",
  "hidden_size": 444,
  "initializer_range": 0.02,
  "intermediate_size": 888,
  "max_position_embeddings": 444,
  "model_type": "gemma",
  "num_attention_heads": 4,
  "num_hidden_layers": 4,
  "num_key_value_heads": 1,
  "pad_token_id": 0,
  "rms_norm_eps": 1e-06,
  "rope_parameters": {
    "rope_theta": 10000.0,
    "rope_type": "default"
  },
  "tie_word_embeddings": true,
  "transformers_version": "5.0.0",
  "use_bidirectional_attention": null,
  "use_cache": false,
  "vocab_size": 8198
}