{ "architectures": ["SixpertMoEForCausalLM"], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151643, "hidden_act": "silu", "hidden_size": 3584, "initializer_range": 0.02, "intermediate_size": 14336, "max_position_embeddings": 131072, "model_type": "sixpert_moe", "num_attention_heads": 28, "num_experts": 16, "num_experts_per_tok": 2, "num_hidden_layers": 28, "num_key_value_heads": 4, "quantization_config": { "quant_method": "gguf", "quantization_type": "Q4_K_M" }, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "float16", "transformers_version": "4.46.0", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936, "model_name": "Sixpert K2", "model_version": "1.0", "context_length": 131072, "authors": ["Inyang David", "Sixtus Matthew"], "license": "apache-2.0", "total_parameters": "~8.9B", "active_parameters": "~1.2B per forward pass" }