{ "measurement": { "model.layers.0": { "accuracy": 0.6830730438232422, "total_bits": 30627648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.1": { "accuracy": 0.7941532135009766, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.2": { "accuracy": 0.7930316925048828, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.3": { "accuracy": 0.8217496871948242, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.4": { "accuracy": 0.8298921585083008, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.5": { "accuracy": 0.7461814880371094, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.6": { "accuracy": 0.9222159385681152, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.7": { "accuracy": 0.9137954711914062, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.8": { "accuracy": 0.9160647392272949, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.9": { "accuracy": 0.9239978790283203, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.10": { "accuracy": 0.9132890701293945, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.11": { "accuracy": 0.908735990524292, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.12": { "accuracy": 0.9123492240905762, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.13": { "accuracy": 0.9198918342590332, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.14": { "accuracy": 0.909644365310669, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.15": { "accuracy": 0.9010679721832275, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.16": { "accuracy": 0.8972649574279785, "total_bits": 28563648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.17": { "accuracy": 0.8895649909973145, "total_bits": 30627648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.18": { "accuracy": 0.9149861335754395, "total_bits": 36303648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.19": { "accuracy": 0.9197125434875488, "total_bits": 36303648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.20": { "accuracy": 0.9247608184814453, "total_bits": 38367648, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.21": { "accuracy": 0.9333062171936035, "total_bits": 41980320, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.22": { "accuracy": 0.9679172039031982, "total_bits": 51784992, "q_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } }, "model.layers.23": { "accuracy": 0.950645923614502, "total_bits": 47656992, "q_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "k_proj": { "group_size": { "2": 64 }, "bits": [ 2 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "v_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "o_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "up_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "gate_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 }, "down_proj": { "group_size": { "4": 128 }, "bits": [ 4 ], "bits_prop": [ 1 ], "scale_bits": 4 } } } }