{ "architectures": [ "Spark2_5ForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "auto_map": { "AutoConfig": "configuration_spark.Spark2_5Config", "AutoModel": "modeling_spark.Spark2_5Model", "AutoModelForCausalLM": "modeling_spark.Spark2_5ForCausalLM" }, "bos_token_id": 0, "dtype": "bfloat16", "eos_token_id": 1, "gate_attn_act_mode": "sigmoid", "head_dim": 256, "headwise_attn_output_gate": true, "hidden_act": "gelu", "hidden_size": 2560, "initializer_range": 0.01976, "intermediate_size": 10240, "layer_types": [ "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention" ], "max_position_embeddings": 1048576, "mlp_bias": false, "model_type": "spark2_5", "num_attention_heads": 16, "num_hidden_layers": 36, "num_key_value_heads": 4, "pad_token_id": 2, "rms_norm_eps": 1e-06, "rope_parameters": { "full_attention": { "partial_rotary_factor": 0.25, "rope_theta": 5000000 }, "sliding_attention": { "partial_rotary_factor": 1.0, "rope_theta": 10000 } }, "sliding_window": 512, "tie_word_embeddings": true, "transformers_version": "4.57.1", "use_cache": true, "vocab_size": 131072 }