| { |
| "_name_or_path": "microsoft/deberta-v3-base", |
| "architectures": [ |
| "DebertaV2ForSequenceClassification" |
| ], |
| "attention_probs_dropout_prob": 0.1, |
| "hidden_act": "gelu", |
| "hidden_dropout_prob": 0.1, |
| "hidden_size": 768, |
| "id2label": { |
| "0": "Incomplete", |
| "1": "Additive", |
| "2": "Duplication", |
| "3": "Wrong_Fraction", |
| "4": "Subtraction", |
| "5": "Positive", |
| "6": "Wrong_term", |
| "7": "Irrelevant", |
| "8": "Inversion", |
| "9": "Mult", |
| "10": "Denominator-only_change", |
| "11": "Whole_numbers_larger", |
| "12": "Adding_across", |
| "13": "WNB", |
| "14": "Tacking", |
| "15": "Unknowable", |
| "16": "SwapDividend", |
| "17": "Scale", |
| "18": "Not_variable", |
| "19": "Firstterm", |
| "20": "Adding_terms", |
| "21": "Multiplying_by_4", |
| "22": "FlipChange", |
| "23": "Division", |
| "24": "Definition", |
| "25": "Interior", |
| "26": "Longer_is_bigger", |
| "27": "Ignores_zeroes", |
| "28": "Shorter_is_bigger", |
| "29": "Base_rate", |
| "30": "Inverse_operation", |
| "31": "Certainty", |
| "32": "Incorrect_equivalent_fraction_addition", |
| "33": "Wrong_Operation" |
| }, |
| "initializer_range": 0.02, |
| "intermediate_size": 3072, |
| "label2id": { |
| "Adding_across": 12, |
| "Adding_terms": 20, |
| "Additive": 1, |
| "Base_rate": 29, |
| "Certainty": 31, |
| "Definition": 24, |
| "Denominator-only_change": 10, |
| "Division": 23, |
| "Duplication": 2, |
| "Firstterm": 19, |
| "FlipChange": 22, |
| "Ignores_zeroes": 27, |
| "Incomplete": 0, |
| "Incorrect_equivalent_fraction_addition": 32, |
| "Interior": 25, |
| "Inverse_operation": 30, |
| "Inversion": 8, |
| "Irrelevant": 7, |
| "Longer_is_bigger": 26, |
| "Mult": 9, |
| "Multiplying_by_4": 21, |
| "Not_variable": 18, |
| "Positive": 5, |
| "Scale": 17, |
| "Shorter_is_bigger": 28, |
| "Subtraction": 4, |
| "SwapDividend": 16, |
| "Tacking": 14, |
| "Unknowable": 15, |
| "WNB": 13, |
| "Whole_numbers_larger": 11, |
| "Wrong_Fraction": 3, |
| "Wrong_Operation": 33, |
| "Wrong_term": 6 |
| }, |
| "layer_norm_eps": 1e-07, |
| "legacy": true, |
| "max_position_embeddings": 512, |
| "max_relative_positions": -1, |
| "model_type": "deberta-v2", |
| "norm_rel_ebd": "layer_norm", |
| "num_attention_heads": 12, |
| "num_hidden_layers": 12, |
| "pad_token_id": 0, |
| "pooler_dropout": 0, |
| "pooler_hidden_act": "gelu", |
| "pooler_hidden_size": 768, |
| "pos_att_type": [ |
| "p2c", |
| "c2p" |
| ], |
| "position_biased_input": false, |
| "position_buckets": 256, |
| "relative_attention": true, |
| "share_att_key": true, |
| "torch_dtype": "float32", |
| "transformers_version": "4.48.0", |
| "type_vocab_size": 0, |
| "vocab_size": 128100 |
| } |
|
|