{ "architectures": [ "T5ForConditionalGeneration" ], "model_type": "t5", "vocab_size": 33728, "d_model": 1024, "d_kv": 64, "d_ff": 2816, "num_layers": 24, "num_decoder_layers": 24, "num_heads": 16, "relative_attention_num_buckets": 32, "relative_attention_max_distance": 128, "dropout_rate": 0.0, "layer_norm_epsilon": 1e-06, "initializer_factor": 1.0, "feed_forward_proj": "gated-gelu", "dense_act_fn": "gelu_new", "is_gated_act": true, "is_encoder_decoder": true, "use_cache": true, "tie_word_embeddings": true, "scale_decoder_outputs": true, "classifier_dropout": 0.0, "pad_token_id": 0, "eos_token_id": 1, "decoder_start_token_id": 0, "max_encoder_length": 2048, "max_decoder_length": 2048 }