{ | |
"_name_or_path": "gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48", | |
"activation": "relu", | |
"architectures": [ | |
"DistilBertForSequenceClassification" | |
], | |
"attention_dropout": 0.05, | |
"cache_dir": null, | |
"dim": 768, | |
"dropout": 0.05, | |
"finetuning_task": "stsb", | |
"hidden_dim": 3072, | |
"id2label": { | |
"0": "LABEL_0" | |
}, | |
"initializer_range": 0.02, | |
"label2id": { | |
"LABEL_0": 0 | |
}, | |
"max_position_embeddings": 512, | |
"model_type": "hybridbert", | |
"n_heads": 12, | |
"n_layers": 12, | |
"pad_token_id": 0, | |
"problem_type": "regression", | |
"qa_dropout": 0.1, | |
"revision": "main", | |
"seq_classif_dropout": 0.1, | |
"sinusoidal_pos_embds": true, | |
"torch_dtype": "float32", | |
"transformers_version": "4.29.2", | |
"use_auth_token": null, | |
"vocab_size": 30522 | |
} | |