ccczy/output_dir
04
1{2 "_name_or_path": "./albert_base_v2_model",3 "architectures": [4 "AlbertForSequenceClassification"5 ],6 "attention_probs_dropout_prob": 0,7 "bos_token_id": 2,8 "classifier_dropout_prob": 0.1,9 "down_scale_factor": 1,10 "embedding_size": 128,11 "eos_token_id": 3,12 "gap_size": 0,13 "hidden_act": "gelu_new",14 "hidden_dropout_prob": 0,15 "hidden_size": 768,16 "id2label": {17 "0": "LABEL_0",18 "1": "LABEL_1",19 "2": "LABEL_2",20 "3": "LABEL_3",21 "4": "LABEL_4",22 "5": "LABEL_5",23 "6": "LABEL_6",24 "7": "LABEL_7",25 "8": "LABEL_8",26 "9": "LABEL_9",27 "10": "LABEL_10",28 "11": "LABEL_11",29 "12": "LABEL_12",30 "13": "LABEL_13",31 "14": "LABEL_14",32 "15": "LABEL_15",33 "16": "LABEL_16",34 "17": "LABEL_17",35 "18": "LABEL_18",36 "19": "LABEL_19",37 "20": "LABEL_20",38 "21": "LABEL_21",39 "22": "LABEL_22",40 "23": "LABEL_23",41 "24": "LABEL_24",42 "25": "LABEL_25",43 "26": "LABEL_26"44 },45 "initializer_range": 0.02,46 "inner_group_num": 1,47 "intermediate_size": 3072,48 "label2id": {49 "LABEL_0": 0,50 "LABEL_1": 1,51 "LABEL_10": 10,52 "LABEL_11": 11,53 "LABEL_12": 12,54 "LABEL_13": 13,55 "LABEL_14": 14,56 "LABEL_15": 15,57 "LABEL_16": 16,58 "LABEL_17": 17,59 "LABEL_18": 18,60 "LABEL_19": 19,61 "LABEL_2": 2,62 "LABEL_20": 20,63 "LABEL_21": 21,64 "LABEL_22": 22,65 "LABEL_23": 23,66 "LABEL_24": 24,67 "LABEL_25": 25,68 "LABEL_26": 26,69 "LABEL_3": 3,70 "LABEL_4": 4,71 "LABEL_5": 5,72 "LABEL_6": 6,73 "LABEL_7": 7,74 "LABEL_8": 8,75 "LABEL_9": 976 },77 "layer_norm_eps": 1e-12,78 "max_position_embeddings": 512,79 "model_type": "albert",80 "net_structure_type": 0,81 "num_attention_heads": 12,82 "num_hidden_groups": 1,83 "num_hidden_layers": 12,84 "num_memory_blocks": 0,85 "pad_token_id": 0,86 "position_embedding_type": "absolute",87 "problem_type": "single_label_classification",88 "torch_dtype": "float32",89 "transformers_version": "4.33.1",90 "type_vocab_size": 2,91 "vocab_size": 3000092}93