{ "_name_or_path": "/data/katsumata/sandbox/megatron/checkpoints/tsubame/hf-release/RetrievaBERT-seq2048-iter454k", "architectures": [ "RetrievaBertForMaskedLM" ], "attention_probs_dropout_prob": 0.1, "auto_map": { "AutoConfig": "configuration_retrieva_bert.RetrievaBertConfig", "AutoModelForMaskedLM": "modeling_retrieva_bert.RetrievaBertForMaskedLM" }, "bos_token_id": 1, "eos_token_id": 2, "hidden_act": "silu", "hidden_dropout_prob": 0.1, "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 4096, "layer_norm_eps": 1e-12, "lm_head_hidden_act": "gelu", "max_position_embeddings": 2048, "max_sequence_length": 2048, "mlp_bias": true, "model_type": "retrieva-bert", "num_attention_heads": 24, "num_hidden_layers": 48, "num_key_value_heads": 1, "pad_token_id": 4, "position_embedding_type": "absolute", "rope_theta": 10000.0, "rotary_percent": 1.0, "torch_dtype": "bfloat16", "transformers_version": "4.41.2", "type_vocab_size": 0, "use_cache": true, "vocab_size": 99584 }