tiiuae/falcon-rw-1b/bert4torch_config.json

{
    "model": "falcon",
    "type_vocab_size": 0,
    "p_bias": "alibi",
    "apply_residual_post_layernorm": false,
    "attention_dropout": 0.0,
    "bias": true,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "hidden_dropout": 0.0,
    "hidden_size": 2048,
    "initializer_range": 0.02,
    "intermediate_size": 8192,
    "hidden_act": "gelu",
    "layer_norm_eps": 1e-05,
    "multi_query": false,
    "num_attention_heads": 32,
    "num_hidden_layers": 24,
    "torch_dtype": "bfloat16",
    "vocab_size": 50304,
    "skip_init": true,
    "norm_mode": "torch_buildin"
    }