bigscience/bloom-560m/bert4torch_config.json

{
    "apply_residual_connection_post_layernorm": false,
    "attention_dropout": 0.0,
    "attention_softmax_in_fp32": true,
    "bias_dropout_fusion": true,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "pad_token_id": 3,
    "unk_token_id": 0,
    "hidden_dropout": 0.0,
    "hidden_act": "gelu_fast",
    "initializer_range": 0.02,
    "layer_norm_eps": 1e-05,
    "hidden_size": 1024,
    "intermediate_size": 4096,
    "num_hidden_layers": 24,
    "num_attention_heads": 16,
    "offset_alibi": 100,
    "pretraining_tp": 1,
    "skip_bias_add": true,
    "skip_bias_add_qkv": false,
    "vocab_size": 250880,
    "segment_vocab_size": 0,
    "pre_layernorm": true,
    "tie_word_embeddings": true,
    "model": "bloom"
  }