| { | |
| "architectures": [ | |
| "CgeForEmbedding" | |
| ], | |
| "attention_dropout": 0.0, | |
| "attn_implementation": "flash_attention_2", | |
| "auto_map": { | |
| "AutoConfig": "configuration_cge2.CGEConfig", | |
| "AutoModel": "modeling_cge2.CgeForEmbedding" | |
| }, | |
| "compressed_dim": 896, | |
| "embedding_method": "pma", | |
| "hidden_act": "silu", | |
| "hidden_size": 896, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4864, | |
| "max_position_embeddings": 32768, | |
| "max_window_layers": 24, | |
| "model_type": "cge2", | |
| "num_attention_heads": 14, | |
| "num_hidden_layers": 24, | |
| "num_key_value_heads": 2, | |
| "padding_side": "left", | |
| "pma_ln": true, | |
| "pma_norm": false, | |
| "pma_norm_mode": "post_normal", | |
| "pma_num_heads": 32, | |
| "rms_norm_eps": 1e-06, | |
| "rope_theta": 1000000.0, | |
| "sliding_window": null, | |
| "tokenizer_name_or_path": "/ainative/muti-modal/yuhang/431428/download_models/CGE2_0B5_1841", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.51.3", | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 151936 | |
| } | |