2020-02-08 03:16:00 +08:00
|
|
|
{
|
|
|
|
"activation": "gelu",
|
|
|
|
"attention_dropout": 0.1,
|
|
|
|
"dim": 768,
|
|
|
|
"dropout": 0.1,
|
|
|
|
"hidden_dim": 3072,
|
|
|
|
"initializer_range": 0.02,
|
|
|
|
"max_position_embeddings": 512,
|
2020-04-24 23:57:51 +08:00
|
|
|
"model_type": "distilbert",
|
2020-02-08 03:16:00 +08:00
|
|
|
"n_heads": 12,
|
|
|
|
"n_layers": 6,
|
|
|
|
"output_past": true,
|
|
|
|
"pad_token_id": 0,
|
|
|
|
"qa_dropout": 0.1,
|
|
|
|
"seq_classif_dropout": 0.2,
|
|
|
|
"sinusoidal_pos_embds": false,
|
|
|
|
"tie_weights_": true,
|
|
|
|
"vocab_size": 28996
|
2020-04-24 23:57:51 +08:00
|
|
|
}
|