jclondonol/Sequence_Classification_Models

1{2  "activation": "gelu",3  "architectures": [4    "DistilBertForSequenceClassification"5  ],6  "attention_dropout": 0.1,7  "bos_token_id": null,8  "dim": 768,9  "dropout": 0.1,10  "dtype": "float32",11  "eos_token_id": null,12  "hidden_dim": 3072,13  "id2label": {14    "0": "math.AC",15    "1": "cs.CV",16    "10": "math.ST",17    "2": "cs.AI",18    "3": "cs.SY",19    "4": "math.GR",20    "5": "cs.CE",21    "6": "cs.PL",22    "7": "cs.IT",23    "8": "cs.DS",24    "9": "cs.NE"25  },26  "initializer_range": 0.02,27  "label2id": {28    "cs.AI": 2,29    "cs.CE": 5,30    "cs.CV": 1,31    "cs.DS": 8,32    "cs.IT": 7,33    "cs.NE": 9,34    "cs.PL": 6,35    "cs.SY": 3,36    "math.AC": 0,37    "math.GR": 4,38    "math.ST": 1039  },40  "max_position_embeddings": 512,41  "model_type": "distilbert",42  "n_heads": 12,43  "n_layers": 6,44  "pad_token_id": 0,45  "problem_type": "single_label_classification",46  "qa_dropout": 0.1,47  "seq_classif_dropout": 0.2,48  "sinusoidal_pos_embds": false,49  "tie_weights_": true,50  "tie_word_embeddings": true,51  "transformers_version": "5.8.1",52  "use_cache": false,53  "vocab_size": 3052254}55