| { | |
| "_name_or_path": "ltg/ltg-bert-babylm", | |
| "architectures": [ | |
| "LtgBertForMaskedLM" | |
| ], | |
| "attention_probs_dropout_prob": 0.1, | |
| "auto_map": { | |
| "AutoConfig": "configuration_ltgbert.LtgBertConfig", | |
| "AutoModelForMaskedLM": "modeling_ltgbert.LtgBertForMaskedLM", | |
| "AutoModelForSequenceClassification": "modeling_ltgbert.LtgBertForSequenceClassification" | |
| }, | |
| "classifier_dropout": 0.2, | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 768, | |
| "intermediate_size": 2048, | |
| "layer_norm_eps": 1e-07, | |
| "max_position_embeddings": 512, | |
| "model_type": "ltgbert", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 12, | |
| "output_all_encoded_layers": true, | |
| "pad_token_id": 4, | |
| "position_bucket_size": 32, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.40.2", | |
| "vocab_size": 16384 | |
| } | |