{ "alibi_starting_size": 1024, "architectures": [ "BertForMaskedLM" ], "auto_map": { "AutoConfig": "Taykhoom/MosaicBERT-updated--configuration_bert.BertConfig", "AutoModel": "Taykhoom/MosaicBERT-updated--bert_layers.BertModel", "AutoModelForMaskedLM": "Taykhoom/MosaicBERT-updated--bert_layers.BertForMaskedLM", "AutoModelForSequenceClassification": "Taykhoom/MosaicBERT-updated--bert_layers.BertForSequenceClassification" }, "attention_probs_dropout_prob": 0.0, "classifier_dropout": null, "dtype": "float32", "hidden_act": "gelu", "hidden_dropout_prob": 0.1, "hidden_size": 768, "initializer_range": 0.02, "intermediate_size": 3072, "layer_norm_eps": 1e-12, "model_max_length": 10000, "model_type": "bert", "num_attention_heads": 12, "num_hidden_layers": 12, "pad_token_id": 0, "transformers_version": "4.57.6", "type_vocab_size": 2, "use_cache": true, "vocab_size": 74 }