{ "architectures": [ "LongT5ForConditionalGeneration" ], "d_ff": 10240, "d_kv": 64, "d_model": 4096, "dense_act_fn": "relu", "dropout_rate": 0.1, "encoder_attention_type": "local", "eos_token_id": 1, "feed_forward_proj": "relu", "global_block_size": 16, "initializer_factor": 1.0, "is_encoder_decoder": true, "is_gated_act": false, "layer_norm_epsilon": 1e-06, "local_radius": 127, "model_type": "longt5", "num_decoder_layers": 48, "num_heads": 64, "num_layers": 48, "pad_token_id": 0, "relative_attention_max_distance": 128, "relative_attention_num_buckets": 32, "torch_dtype": "float32", "transformers_version": "4.25.0.dev0", "use_cache": true, "vocab_size": 32128 }