KalaiselvanD/mega_08_1

1{2  "_name_or_path": "mnaylor/mega-base-wikitext",3  "activation": "silu",4  "add_lm_hidden_dense_layer": false,5  "add_token_type_embeddings": false,6  "architectures": [7    "MegaForSequenceClassification"8  ],9  "attention_activation": "softmax",10  "attention_probs_dropout_prob": 0.1,11  "bidirectional": true,12  "bos_token_id": 0,13  "chunk_size": -1,14  "classifier_dropout": null,15  "dropout_prob": 0.1,16  "ema_beta_range": 0.02,17  "ema_delta_alpha_range": 0.2,18  "ema_gamma_omega_range": 1.0,19  "ema_projection_size": 16,20  "eos_token_id": 2,21  "hidden_dropout_prob": 0.1,22  "hidden_size": 128,23  "id2label": {24    "0": "low",25    "1": "medium",26    "2": "high"27  },28  "initializer_range": 0.02,29  "intermediate_size": 256,30  "label2id": {31    "high": 2,32    "low": 0,33    "medium": 134  },35  "max_positions": 10000,36  "model_type": "mega",37  "nffn_activation_dropout_prob": 0.0,38  "nffn_hidden_size": 256,39  "norm_affine": true,40  "normalization_type": "scalenorm",41  "normalize_before_ffn": true,42  "normalize_before_mega": true,43  "num_attention_heads": 1,44  "num_hidden_layers": 4,45  "pad_token_id": 1,46  "problem_type": "single_label_classification",47  "relative_positional_bias": "rotary",48  "shared_representation_size": 64,49  "torch_dtype": "float32",50  "transformers_version": "4.39.3",51  "truncation": null,52  "type_vocab_size": 2,53  "use_cache": true,54  "use_chunking": false,55  "use_feature_dropout": false,56  "use_normalized_ffn": true,57  "vocab_size": 5026558}59