{ "_sliding_window_pattern": 6, "architectures": [ "GearForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "attn_logit_softcapping": null, "auto_map": { "AutoConfig": "configuration_gear.GearConfig", "AutoModelForCausalLM": "modeling_gear.GearForCausalLM" }, "bos_token_id": 1, "conv_L_cache": 4, "dtype": "bfloat16", "eos_token_id": 2, "final_logit_softcapping": null, "head_dim": 64, "hidden_activation": "relu6", "hidden_size": 128, "initializer_range": 0.02, "intermediate_size": 272, "layer_types": [ "conv_mixer", "conv_mixer", "full_attention", "conv_mixer", "conv_mixer" ], "max_position_embeddings": 512, "model_type": "gear", "num_attention_heads": 2, "num_hidden_layers": 5, "num_key_value_heads": 1, "pad_token_id": 0, "query_pre_attn_scalar": 64, "rms_norm_eps": 1e-06, "rope_local_base_freq": 10000, "rope_scaling": null, "rope_theta": 1000000, "sliding_window": 512, "transformers_version": "4.57.3", "use_bidirectional_attention": false, "use_cache": true, "vocab_size": 2048 }