visolex/bartpho-hsd-span
053
1{2 "activation_dropout": 0.0,3 "activation_function": "gelu",4 "architectures": [5 "T5EncoderModel"6 ],7 "attention_dropout": 0.0,8 "bos_token_id": 0,9 "classifier_dropout": 0.0,10 "d_ff": 2048,11 "d_kv": 64,12 "d_model": 1024,13 "decoder_attention_heads": 16,14 "decoder_ffn_dim": 4096,15 "decoder_layerdrop": 0.0,16 "decoder_layers": 12,17 "decoder_start_token_id": 2,18 "dense_act_fn": "relu",19 "dropout": 0.1,20 "dropout_rate": 0.1,21 "dtype": "float32",22 "encoder_attention_heads": 16,23 "encoder_ffn_dim": 4096,24 "encoder_layerdrop": 0.0,25 "encoder_layers": 12,26 "eos_token_id": 2,27 "feed_forward_proj": "relu",28 "forced_eos_token_id": 2,29 "gradient_checkpointing": false,30 "init_std": 0.02,31 "initializer_factor": 1.0,32 "is_encoder_decoder": false,33 "is_gated_act": false,34 "layer_norm_epsilon": 1e-06,35 "max_position_embeddings": 1024,36 "model_type": "t5",37 "num_decoder_layers": 6,38 "num_heads": 8,39 "num_layers": 12,40 "pad_token_id": 1,41 "relative_attention_max_distance": 128,42 "relative_attention_num_buckets": 32,43 "scale_embedding": false,44 "tokenizer_class": "BartphoTokenizer",45 "transformers_version": "4.57.1",46 "use_cache": false,47 "vocab_size": 4003048}49 