{ "activation_dropout": 0.0, "architectures": [ "FunnelModel" ], "attention_dropout": 0.1, "attention_type": "relative_shift", "block_repeats": [ 1, 1, 1 ], "block_sizes": [ 6, 6, 6 ], "d_head": 64, "d_inner": 3072, "d_model": 768, "embedding_size": 768, "hidden_act": "gelu_new", "hidden_dropout": 0.1, "initializer_range": 0.1, "initializer_std": 0.02, "layer_norm_eps": 1e-09, "max_position_embeddings": 512, "model_type": "funnel", "n_head": 12, "num_decoder_layers": 2, "pool_q_only": true, "pooling_type": "mean", "rel_attn_type": "factorized", "separate_cls": true, "truncate_seq": true, "type_vocab_size": 3, "vocab_size": 42000 }