bert-large-japanese / config.json
singletongue's picture
First commit of the bert-large-japanese model and tokenizer.
2cbf66b
raw
history blame
336 Bytes
{
"attention_probs_dropout_prob": 0.1,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 4096,
"max_position_embeddings": 512,
"num_attention_heads": 16,
"num_hidden_layers": 24,
"type_vocab_size": 2,
"vocab_size": 32768
}