{ "_name_or_path": "MIRA", "apply_aux_loss": true, "architectures": [ "MIRAForPrediction" ], "attention_dropout": 0.0, "gradient_checkpointing_kwargs": { "use_reentrant": true }, "hidden_act": "silu", "hidden_size": 768, "horizon_lengths": [ 1 ], "initializer_range": 0.02, "input_size": 1, "intermediate_size": 3072, "max_position_embeddings": 4096, "model_type": "mira", "num_attention_heads": 12, "num_experts": 8, "num_experts_per_tok": 2, "num_hidden_layers": 12, "num_key_value_heads": 12, "ode_activation_threshold": 1.0, "ode_func_activation": "silu", "ode_func_hidden_dims": [ 128, 128 ], "ode_func_use_time": true, "ode_solver_atol": 1e-06, "ode_solver_method": "dopri5", "ode_solver_rtol": 1e-06, "rms_norm_eps": 1e-06, "rope_theta": 10000, "router_aux_loss_factor": 0.02, "tie_word_embeddings": false, "time_aware_rotary": true, "time_scale": 1.0, "torch_dtype": "bfloat16", "transformers_version": "4.40.1", "use_cache": true, "use_dense": false, "use_terminal_ode": true }