config.json

{
  "architectures": [
    "LlamaForCausalLM"
  ],
  "bos_token_id": null,
  "eos_token_id": 0,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 13312,
  "max_position_embeddings": 131072,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 28,
  "num_key_value_heads": 8,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "factor": 16.0,
    "high_freq_factor": 4.0,
    "low_freq_factor": 1.0,
    "original_max_position_embeddings": 8192,
    "rope_type": "llama3"
  },
  "rope_theta": 500000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "use_cache": true,
  "vocab_size": 65536
}