| { | |
| "architectures": [ | |
| "MariaForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_maria.MariaConfig", | |
| "AutoModelForCausalLM": "modeling_maria.MariaForCausalLM" | |
| }, | |
| "model_type": "maria", | |
| "vocab_size": 128256, | |
| "hidden_size": 2048, | |
| "intermediate_size": 5632, | |
| "num_hidden_layers": 18, | |
| "num_attention_heads": 16, | |
| "num_key_value_heads": 4, | |
| "max_position_embeddings": 8192, | |
| "rope_theta": 500000.0, | |
| "attention_bias": false, | |
| "layer_norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_weights": true, | |
| "bos_token_id": 128000, | |
| "eos_token_id": 128009, | |
| "pad_token_id": 128004, | |
| "torch_dtype": "float16", | |
| "tie_word_embeddings": true | |
| } |