{ "architectures": [ "ZayaForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 2, "cca_time0": 2, "cca_time1": 2, "dtype": "float32", "eos_token_id": 106, "head_dim": 16, "hidden_act": "gelu", "hidden_size": 32, "initializer_range": 0.02, "is_decoder": true, "layer_types": [ "hybrid", "hybrid_sliding" ], "lm_head_bias": false, "max_position_embeddings": 512, "model_type": "zaya", "moe_intermediate_size": 32, "num_attention_heads": 2, "num_experts": 8, "num_experts_per_tok": 1, "num_hidden_layers": 2, "num_key_value_heads": 2, "output_router_logits": false, "pad_token_id": 0, "rms_norm_eps": 1e-05, "rope_parameters": { "hybrid": { "partial_rotary_factor": 0.5, "rope_theta": 5000000.0, "rope_type": "default" }, "hybrid_sliding": { "partial_rotary_factor": 0.5, "rope_theta": 10000.0, "rope_type": "default" } }, "router_hidden_size": 256, "sliding_window": 64, "tie_word_embeddings": false, "transformers_version": "5.16.0.dev0", "use_cache": true, "vocab_size": 262147 }