{ "architectures": [ "SolarOpen2ForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "auto_map": { "AutoConfig": "configuration_solar_open2.SolarOpen2Config", "AutoModel": "modeling_solar_open2.SolarOpen2Model", "AutoModelForCausalLM": "modeling_solar_open2.SolarOpen2ForCausalLM" }, "bos_token_id": 1, "dtype": "bfloat16", "eos_token_id": 2, "first_k_dense_replace": 0, "gqa_interval": 3, "gqa_layers": [ 0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44 ], "hc_dynamic": true, "hc_rate": 0, "hc_use_sigmoid": false, "hc_use_tanh": true, "head_dim": 128, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 10240, "kda_allow_neg_eigval": true, "kda_gate_lower_bound": -5.0, "kda_use_full_proj": false, "linear_attn_config": { "head_dim": 128, "num_heads": 64, "num_kv_heads": null, "short_conv_kernel_size": 4 }, "max_position_embeddings": 1048576, "model_type": "solar_open2", "moe_intermediate_size": 1280, "n_group": 1, "n_routed_experts": 320, "n_shared_experts": 1, "norm_topk_prob": true, "num_attention_heads": 64, "num_experts_per_tok": 8, "num_hidden_layers": 48, "num_key_value_heads": 8, "pad_token_id": 2, "partial_rotary_factor": 1.0, "quantization_config": { "bits": 4, "block_name_to_quantize": "model.layers", "data_type": "int", "extra_config": { ".*model\\.layers\\.[0-47]\\.mlp\\.gate.*": { "bits": 16, "data_type": "float" }, "model.layers.0.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.0.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.1.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.10.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.11.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.12.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.13.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.14.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.15.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.16.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.17.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.18.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.19.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.2.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.20.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.21.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.22.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.23.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.24.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.25.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.26.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.27.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.28.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.29.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.3.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.30.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.31.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.32.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.33.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.34.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.35.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.36.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.37.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.38.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.39.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.4.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.40.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.41.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.42.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.43.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.44.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.45.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.46.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.47.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.5.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.6.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.7.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.self_attn.g_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.8.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.mlp.shared_experts.down_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.mlp.shared_experts.gate_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.mlp.shared_experts.up_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.f_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.f_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.g_a_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.g_b_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.k_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.o_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.q_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" }, "model.layers.9.self_attn.v_proj": { "act_data_type": "fp", "bits": 16, "data_type": "fp" } }, "group_size": 128, "low_gpu_mem_usage": true, "packing_format": "auto_round:auto_gptq", "quant_method": "auto-round", "sym": true }, "rms_norm_eps": 1e-05, "rope_scaling": null, "rope_theta": 10000, "routed_scaling_factor": 1.0, "tie_word_embeddings": false, "topk_group": 1, "transformers_version": "4.57.6", "use_cache": true, "use_gqa_gate": true, "use_gqa_gate_bias": false, "use_qk_norm": false, "use_rope": false, "vocab_size": 196608 }