{ "0": { "layer": 0, "is_moe": true, "routed_experts": { "median_var64": 0.0575, "q25_var64": 0.0572, "q75_var64": 0.058, "mean_s0": 4.01, "std_s0": 0.22, "mean_s0_ratio": 1.03, "n_experts": 384 }, "shared_expert": { "var64": 0.0977, "s0": 6.0, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.39, "s0": 9.75, "s0_s1": 1.46, "shape": [ 384, 7168 ] } }, "1": { "layer": 1, "is_moe": true, "routed_experts": { "median_var64": 0.0568, "q25_var64": 0.0566, "q75_var64": 0.0571, "mean_s0": 3.88, "std_s0": 0.13, "mean_s0_ratio": 1.01, "n_experts": 384 }, "shared_expert": { "var64": 0.104, "s0": 5.85, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3943, "s0": 10.02, "s0_s1": 1.73, "shape": [ 384, 7168 ] } }, "2": { "layer": 2, "is_moe": true, "routed_experts": { "median_var64": 0.0575, "q25_var64": 0.0574, "q75_var64": 0.0576, "mean_s0": 3.84, "std_s0": 0.14, "mean_s0_ratio": 1.01, "n_experts": 384 }, "shared_expert": { "var64": 0.1357, "s0": 6.01, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.2797, "s0": 5.36, "s0_s1": 1.05, "shape": [ 384, 7168 ] } }, "3": { "layer": 3, "is_moe": true, "routed_experts": { "median_var64": 0.0632, "q25_var64": 0.0605, "q75_var64": 0.0653, "mean_s0": 4.66, "std_s0": 0.61, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.0848, "s0": 6.55, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.382, "s0": 8.4, "s0_s1": 1.17, "shape": [ 384, 7168 ] } }, "4": { "layer": 4, "is_moe": true, "routed_experts": { "median_var64": 0.0621, "q25_var64": 0.0604, "q75_var64": 0.065, "mean_s0": 4.57, "std_s0": 0.43, "mean_s0_ratio": 1.07, "n_experts": 384 }, "shared_expert": { "var64": 0.0909, "s0": 5.44, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3823, "s0": 9.06, "s0_s1": 1.28, "shape": [ 384, 7168 ] } }, "5": { "layer": 5, "is_moe": true, "routed_experts": { "median_var64": 0.0634, "q25_var64": 0.0612, "q75_var64": 0.0662, "mean_s0": 4.75, "std_s0": 0.61, "mean_s0_ratio": 1.11, "n_experts": 384 }, "shared_expert": { "var64": 0.087, "s0": 5.68, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3836, "s0": 9.16, "s0_s1": 1.42, "shape": [ 384, 7168 ] } }, "6": { "layer": 6, "is_moe": true, "routed_experts": { "median_var64": 0.063, "q25_var64": 0.0612, "q75_var64": 0.0657, "mean_s0": 4.78, "std_s0": 0.54, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1026, "s0": 6.42, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3454, "s0": 7.85, "s0_s1": 1.24, "shape": [ 384, 7168 ] } }, "7": { "layer": 7, "is_moe": true, "routed_experts": { "median_var64": 0.0612, "q25_var64": 0.059, "q75_var64": 0.0634, "mean_s0": 4.6, "std_s0": 0.55, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.0995, "s0": 5.87, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3136, "s0": 8.55, "s0_s1": 1.3, "shape": [ 384, 7168 ] } }, "8": { "layer": 8, "is_moe": true, "routed_experts": { "median_var64": 0.0623, "q25_var64": 0.06, "q75_var64": 0.0654, "mean_s0": 4.75, "std_s0": 0.63, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1021, "s0": 6.26, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3191, "s0": 8.17, "s0_s1": 1.27, "shape": [ 384, 7168 ] } }, "9": { "layer": 9, "is_moe": true, "routed_experts": { "median_var64": 0.0625, "q25_var64": 0.0594, "q75_var64": 0.0659, "mean_s0": 4.84, "std_s0": 0.64, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.0912, "s0": 5.31, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3433, "s0": 8.0, "s0_s1": 1.27, "shape": [ 384, 7168 ] } }, "10": { "layer": 10, "is_moe": true, "routed_experts": { "median_var64": 0.0634, "q25_var64": 0.0607, "q75_var64": 0.0666, "mean_s0": 4.91, "std_s0": 0.65, "mean_s0_ratio": 1.13, "n_experts": 384 }, "shared_expert": { "var64": 0.0986, "s0": 5.7, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3424, "s0": 7.67, "s0_s1": 1.19, "shape": [ 384, 7168 ] } }, "11": { "layer": 11, "is_moe": true, "routed_experts": { "median_var64": 0.0642, "q25_var64": 0.0613, "q75_var64": 0.068, "mean_s0": 5.15, "std_s0": 0.79, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1124, "s0": 6.18, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3379, "s0": 8.29, "s0_s1": 1.31, "shape": [ 384, 7168 ] } }, "12": { "layer": 12, "is_moe": true, "routed_experts": { "median_var64": 0.066, "q25_var64": 0.0629, "q75_var64": 0.0703, "mean_s0": 5.26, "std_s0": 0.83, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.116, "s0": 6.29, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.349, "s0": 8.95, "s0_s1": 1.43, "shape": [ 384, 7168 ] } }, "13": { "layer": 13, "is_moe": true, "routed_experts": { "median_var64": 0.0662, "q25_var64": 0.0631, "q75_var64": 0.0701, "mean_s0": 5.28, "std_s0": 0.88, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1161, "s0": 6.79, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3374, "s0": 8.56, "s0_s1": 1.23, "shape": [ 384, 7168 ] } }, "14": { "layer": 14, "is_moe": true, "routed_experts": { "median_var64": 0.0673, "q25_var64": 0.063, "q75_var64": 0.0744, "mean_s0": 5.41, "std_s0": 1.06, "mean_s0_ratio": 1.15, "n_experts": 384 }, "shared_expert": { "var64": 0.118, "s0": 6.98, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3822, "s0": 10.46, "s0_s1": 1.3, "shape": [ 384, 7168 ] } }, "15": { "layer": 15, "is_moe": true, "routed_experts": { "median_var64": 0.0692, "q25_var64": 0.0639, "q75_var64": 0.0742, "mean_s0": 5.46, "std_s0": 0.96, "mean_s0_ratio": 1.18, "n_experts": 384 }, "shared_expert": { "var64": 0.1134, "s0": 6.65, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3733, "s0": 9.9, "s0_s1": 1.38, "shape": [ 384, 7168 ] } }, "16": { "layer": 16, "is_moe": true, "routed_experts": { "median_var64": 0.0711, "q25_var64": 0.0655, "q75_var64": 0.0771, "mean_s0": 5.55, "std_s0": 0.93, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1292, "s0": 6.7, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3777, "s0": 10.55, "s0_s1": 1.6, "shape": [ 384, 7168 ] } }, "17": { "layer": 17, "is_moe": true, "routed_experts": { "median_var64": 0.0754, "q25_var64": 0.068, "q75_var64": 0.0832, "mean_s0": 5.91, "std_s0": 1.08, "mean_s0_ratio": 1.18, "n_experts": 384 }, "shared_expert": { "var64": 0.1343, "s0": 6.87, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3727, "s0": 10.49, "s0_s1": 1.4, "shape": [ 384, 7168 ] } }, "18": { "layer": 18, "is_moe": true, "routed_experts": { "median_var64": 0.0778, "q25_var64": 0.0708, "q75_var64": 0.0873, "mean_s0": 6.16, "std_s0": 1.11, "mean_s0_ratio": 1.19, "n_experts": 384 }, "shared_expert": { "var64": 0.1263, "s0": 6.47, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3929, "s0": 10.48, "s0_s1": 1.27, "shape": [ 384, 7168 ] } }, "19": { "layer": 19, "is_moe": true, "routed_experts": { "median_var64": 0.0831, "q25_var64": 0.0722, "q75_var64": 0.0949, "mean_s0": 6.48, "std_s0": 1.32, "mean_s0_ratio": 1.2, "n_experts": 384 }, "shared_expert": { "var64": 0.1325, "s0": 6.95, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.38, "s0": 11.24, "s0_s1": 1.59, "shape": [ 384, 7168 ] } }, "20": { "layer": 20, "is_moe": true, "routed_experts": { "median_var64": 0.0862, "q25_var64": 0.075, "q75_var64": 0.0993, "mean_s0": 6.81, "std_s0": 1.48, "mean_s0_ratio": 1.19, "n_experts": 384 }, "shared_expert": { "var64": 0.138, "s0": 7.0, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3961, "s0": 10.64, "s0_s1": 1.45, "shape": [ 384, 7168 ] } }, "21": { "layer": 21, "is_moe": true, "routed_experts": { "median_var64": 0.0832, "q25_var64": 0.0748, "q75_var64": 0.0944, "mean_s0": 6.83, "std_s0": 1.4, "mean_s0_ratio": 1.22, "n_experts": 384 }, "shared_expert": { "var64": 0.1391, "s0": 7.84, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3614, "s0": 10.51, "s0_s1": 1.35, "shape": [ 384, 7168 ] } }, "22": { "layer": 22, "is_moe": true, "routed_experts": { "median_var64": 0.0938, "q25_var64": 0.0814, "q75_var64": 0.106, "mean_s0": 7.18, "std_s0": 1.34, "mean_s0_ratio": 1.19, "n_experts": 384 }, "shared_expert": { "var64": 0.1501, "s0": 7.57, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3905, "s0": 10.79, "s0_s1": 1.55, "shape": [ 384, 7168 ] } }, "23": { "layer": 23, "is_moe": true, "routed_experts": { "median_var64": 0.0824, "q25_var64": 0.0741, "q75_var64": 0.0921, "mean_s0": 6.83, "std_s0": 1.28, "mean_s0_ratio": 1.21, "n_experts": 384 }, "shared_expert": { "var64": 0.147, "s0": 7.68, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3442, "s0": 8.23, "s0_s1": 1.09, "shape": [ 384, 7168 ] } }, "24": { "layer": 24, "is_moe": true, "routed_experts": { "median_var64": 0.0875, "q25_var64": 0.0754, "q75_var64": 0.0985, "mean_s0": 7.02, "std_s0": 1.42, "mean_s0_ratio": 1.19, "n_experts": 384 }, "shared_expert": { "var64": 0.1703, "s0": 9.57, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3494, "s0": 9.62, "s0_s1": 1.35, "shape": [ 384, 7168 ] } }, "25": { "layer": 25, "is_moe": true, "routed_experts": { "median_var64": 0.0713, "q25_var64": 0.0647, "q75_var64": 0.0827, "mean_s0": 6.08, "std_s0": 1.44, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1515, "s0": 7.85, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.329, "s0": 8.72, "s0_s1": 1.2, "shape": [ 384, 7168 ] } }, "26": { "layer": 26, "is_moe": true, "routed_experts": { "median_var64": 0.0775, "q25_var64": 0.0667, "q75_var64": 0.0951, "mean_s0": 6.49, "std_s0": 1.58, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1703, "s0": 8.86, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3624, "s0": 11.72, "s0_s1": 1.56, "shape": [ 384, 7168 ] } }, "27": { "layer": 27, "is_moe": true, "routed_experts": { "median_var64": 0.0633, "q25_var64": 0.0583, "q75_var64": 0.07, "mean_s0": 4.95, "std_s0": 1.0, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1167, "s0": 7.37, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3594, "s0": 11.8, "s0_s1": 1.26, "shape": [ 384, 7168 ] } }, "28": { "layer": 28, "is_moe": true, "routed_experts": { "median_var64": 0.0713, "q25_var64": 0.0632, "q75_var64": 0.083, "mean_s0": 5.75, "std_s0": 1.4, "mean_s0_ratio": 1.14, "n_experts": 384 }, "shared_expert": { "var64": 0.1461, "s0": 8.75, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3393, "s0": 10.69, "s0_s1": 1.25, "shape": [ 384, 7168 ] } }, "29": { "layer": 29, "is_moe": true, "routed_experts": { "median_var64": 0.0632, "q25_var64": 0.0604, "q75_var64": 0.0726, "mean_s0": 5.29, "std_s0": 1.66, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.1844, "s0": 8.93, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3366, "s0": 11.12, "s0_s1": 1.25, "shape": [ 384, 7168 ] } }, "30": { "layer": 30, "is_moe": true, "routed_experts": { "median_var64": 0.0704, "q25_var64": 0.0622, "q75_var64": 0.0853, "mean_s0": 5.9, "std_s0": 1.49, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1669, "s0": 8.9, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.341, "s0": 10.78, "s0_s1": 1.21, "shape": [ 384, 7168 ] } }, "31": { "layer": 31, "is_moe": true, "routed_experts": { "median_var64": 0.062, "q25_var64": 0.0601, "q75_var64": 0.0668, "mean_s0": 5.08, "std_s0": 1.85, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.2215, "s0": 11.11, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3385, "s0": 11.69, "s0_s1": 1.27, "shape": [ 384, 7168 ] } }, "32": { "layer": 32, "is_moe": true, "routed_experts": { "median_var64": 0.0667, "q25_var64": 0.0624, "q75_var64": 0.0806, "mean_s0": 5.86, "std_s0": 1.98, "mean_s0_ratio": 1.17, "n_experts": 384 }, "shared_expert": { "var64": 0.2033, "s0": 8.86, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3422, "s0": 11.73, "s0_s1": 1.2, "shape": [ 384, 7168 ] } }, "33": { "layer": 33, "is_moe": true, "routed_experts": { "median_var64": 0.0627, "q25_var64": 0.0585, "q75_var64": 0.0772, "mean_s0": 5.78, "std_s0": 3.5, "mean_s0_ratio": 1.14, "n_experts": 384 }, "shared_expert": { "var64": 0.2168, "s0": 9.23, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3471, "s0": 11.29, "s0_s1": 1.21, "shape": [ 384, 7168 ] } }, "34": { "layer": 34, "is_moe": true, "routed_experts": { "median_var64": 0.0643, "q25_var64": 0.0595, "q75_var64": 0.0802, "mean_s0": 5.88, "std_s0": 1.94, "mean_s0_ratio": 1.17, "n_experts": 384 }, "shared_expert": { "var64": 0.1887, "s0": 9.69, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3414, "s0": 9.73, "s0_s1": 1.13, "shape": [ 384, 7168 ] } }, "35": { "layer": 35, "is_moe": true, "routed_experts": { "median_var64": 0.0595, "q25_var64": 0.0576, "q75_var64": 0.0677, "mean_s0": 5.04, "std_s0": 3.29, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.1243, "s0": 7.12, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.338, "s0": 9.81, "s0_s1": 1.2, "shape": [ 384, 7168 ] } }, "36": { "layer": 36, "is_moe": true, "routed_experts": { "median_var64": 0.0611, "q25_var64": 0.0585, "q75_var64": 0.0681, "mean_s0": 5.38, "std_s0": 1.57, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1628, "s0": 7.87, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3288, "s0": 9.61, "s0_s1": 1.1, "shape": [ 384, 7168 ] } }, "37": { "layer": 37, "is_moe": true, "routed_experts": { "median_var64": 0.0613, "q25_var64": 0.0588, "q75_var64": 0.0674, "mean_s0": 5.46, "std_s0": 1.54, "mean_s0_ratio": 1.2, "n_experts": 384 }, "shared_expert": { "var64": 0.1941, "s0": 9.34, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3293, "s0": 10.33, "s0_s1": 1.15, "shape": [ 384, 7168 ] } }, "38": { "layer": 38, "is_moe": true, "routed_experts": { "median_var64": 0.0601, "q25_var64": 0.0583, "q75_var64": 0.0654, "mean_s0": 5.02, "std_s0": 1.31, "mean_s0_ratio": 1.14, "n_experts": 384 }, "shared_expert": { "var64": 0.1466, "s0": 7.15, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3148, "s0": 8.74, "s0_s1": 1.07, "shape": [ 384, 7168 ] } }, "39": { "layer": 39, "is_moe": true, "routed_experts": { "median_var64": 0.0586, "q25_var64": 0.0577, "q75_var64": 0.062, "mean_s0": 4.72, "std_s0": 1.37, "mean_s0_ratio": 1.13, "n_experts": 384 }, "shared_expert": { "var64": 0.1799, "s0": 11.33, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3228, "s0": 9.4, "s0_s1": 1.11, "shape": [ 384, 7168 ] } }, "40": { "layer": 40, "is_moe": true, "routed_experts": { "median_var64": 0.059, "q25_var64": 0.0578, "q75_var64": 0.0622, "mean_s0": 4.71, "std_s0": 1.18, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.1575, "s0": 9.11, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3221, "s0": 8.06, "s0_s1": 1.05, "shape": [ 384, 7168 ] } }, "41": { "layer": 41, "is_moe": true, "routed_experts": { "median_var64": 0.059, "q25_var64": 0.0579, "q75_var64": 0.0615, "mean_s0": 4.69, "std_s0": 1.45, "mean_s0_ratio": 1.11, "n_experts": 384 }, "shared_expert": { "var64": 0.1898, "s0": 9.29, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3202, "s0": 8.5, "s0_s1": 1.07, "shape": [ 384, 7168 ] } }, "42": { "layer": 42, "is_moe": true, "routed_experts": { "median_var64": 0.059, "q25_var64": 0.0577, "q75_var64": 0.0618, "mean_s0": 4.64, "std_s0": 1.23, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1411, "s0": 7.29, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3183, "s0": 7.87, "s0_s1": 1.13, "shape": [ 384, 7168 ] } }, "43": { "layer": 43, "is_moe": true, "routed_experts": { "median_var64": 0.0592, "q25_var64": 0.0581, "q75_var64": 0.0624, "mean_s0": 4.59, "std_s0": 1.35, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.1938, "s0": 11.23, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3211, "s0": 8.92, "s0_s1": 1.18, "shape": [ 384, 7168 ] } }, "44": { "layer": 44, "is_moe": true, "routed_experts": { "median_var64": 0.0599, "q25_var64": 0.0583, "q75_var64": 0.0637, "mean_s0": 4.73, "std_s0": 1.33, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.1642, "s0": 7.72, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.307, "s0": 7.73, "s0_s1": 1.14, "shape": [ 384, 7168 ] } }, "45": { "layer": 45, "is_moe": true, "routed_experts": { "median_var64": 0.0599, "q25_var64": 0.0587, "q75_var64": 0.0626, "mean_s0": 4.59, "std_s0": 1.28, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.2126, "s0": 11.41, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3176, "s0": 8.54, "s0_s1": 1.14, "shape": [ 384, 7168 ] } }, "46": { "layer": 46, "is_moe": true, "routed_experts": { "median_var64": 0.0596, "q25_var64": 0.0583, "q75_var64": 0.0623, "mean_s0": 4.54, "std_s0": 1.19, "mean_s0_ratio": 1.08, "n_experts": 384 }, "shared_expert": { "var64": 0.147, "s0": 7.63, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3127, "s0": 7.96, "s0_s1": 1.2, "shape": [ 384, 7168 ] } }, "47": { "layer": 47, "is_moe": true, "routed_experts": { "median_var64": 0.0597, "q25_var64": 0.0586, "q75_var64": 0.0629, "mean_s0": 4.52, "std_s0": 1.22, "mean_s0_ratio": 1.07, "n_experts": 384 }, "shared_expert": { "var64": 0.1935, "s0": 11.01, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3164, "s0": 8.31, "s0_s1": 1.1, "shape": [ 384, 7168 ] } }, "48": { "layer": 48, "is_moe": true, "routed_experts": { "median_var64": 0.0609, "q25_var64": 0.0593, "q75_var64": 0.0648, "mean_s0": 4.72, "std_s0": 1.29, "mean_s0_ratio": 1.09, "n_experts": 384 }, "shared_expert": { "var64": 0.2037, "s0": 10.5, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3108, "s0": 7.81, "s0_s1": 1.12, "shape": [ 384, 7168 ] } }, "49": { "layer": 49, "is_moe": true, "routed_experts": { "median_var64": 0.0625, "q25_var64": 0.0598, "q75_var64": 0.0678, "mean_s0": 5.17, "std_s0": 1.48, "mean_s0_ratio": 1.12, "n_experts": 384 }, "shared_expert": { "var64": 0.229, "s0": 11.69, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.315, "s0": 8.05, "s0_s1": 1.14, "shape": [ 384, 7168 ] } }, "50": { "layer": 50, "is_moe": true, "routed_experts": { "median_var64": 0.0595, "q25_var64": 0.0581, "q75_var64": 0.0635, "mean_s0": 4.79, "std_s0": 1.37, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1906, "s0": 11.75, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3197, "s0": 7.99, "s0_s1": 1.23, "shape": [ 384, 7168 ] } }, "51": { "layer": 51, "is_moe": true, "routed_experts": { "median_var64": 0.0587, "q25_var64": 0.0577, "q75_var64": 0.0619, "mean_s0": 4.6, "std_s0": 1.43, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1638, "s0": 8.77, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3368, "s0": 9.16, "s0_s1": 1.14, "shape": [ 384, 7168 ] } }, "52": { "layer": 52, "is_moe": true, "routed_experts": { "median_var64": 0.0597, "q25_var64": 0.0581, "q75_var64": 0.0649, "mean_s0": 4.85, "std_s0": 1.47, "mean_s0_ratio": 1.1, "n_experts": 384 }, "shared_expert": { "var64": 0.1783, "s0": 8.39, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3228, "s0": 7.55, "s0_s1": 1.14, "shape": [ 384, 7168 ] } }, "53": { "layer": 53, "is_moe": true, "routed_experts": { "median_var64": 0.07, "q25_var64": 0.0644, "q75_var64": 0.0808, "mean_s0": 6.25, "std_s0": 2.09, "mean_s0_ratio": 1.18, "n_experts": 384 }, "shared_expert": { "var64": 0.1736, "s0": 12.55, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3416, "s0": 7.0, "s0_s1": 1.18, "shape": [ 384, 7168 ] } }, "54": { "layer": 54, "is_moe": true, "routed_experts": { "median_var64": 0.064, "q25_var64": 0.0597, "q75_var64": 0.0711, "mean_s0": 5.76, "std_s0": 2.09, "mean_s0_ratio": 1.16, "n_experts": 384 }, "shared_expert": { "var64": 0.1866, "s0": 12.93, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3352, "s0": 6.83, "s0_s1": 1.05, "shape": [ 384, 7168 ] } }, "55": { "layer": 55, "is_moe": true, "routed_experts": { "median_var64": 0.0623, "q25_var64": 0.0593, "q75_var64": 0.0715, "mean_s0": 5.75, "std_s0": 2.18, "mean_s0_ratio": 1.14, "n_experts": 384 }, "shared_expert": { "var64": 0.1731, "s0": 10.28, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3291, "s0": 6.61, "s0_s1": 1.06, "shape": [ 384, 7168 ] } }, "56": { "layer": 56, "is_moe": true, "routed_experts": { "median_var64": 0.0637, "q25_var64": 0.06, "q75_var64": 0.0722, "mean_s0": 5.85, "std_s0": 2.21, "mean_s0_ratio": 1.15, "n_experts": 384 }, "shared_expert": { "var64": 0.1649, "s0": 10.88, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3432, "s0": 6.87, "s0_s1": 1.18, "shape": [ 384, 7168 ] } }, "57": { "layer": 57, "is_moe": true, "routed_experts": { "median_var64": 0.0746, "q25_var64": 0.0676, "q75_var64": 0.0863, "mean_s0": 7.73, "std_s0": 3.57, "mean_s0_ratio": 1.23, "n_experts": 384 }, "shared_expert": { "var64": 0.1796, "s0": 8.96, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3453, "s0": 5.65, "s0_s1": 1.07, "shape": [ 384, 7168 ] } }, "58": { "layer": 58, "is_moe": true, "routed_experts": { "median_var64": 0.0663, "q25_var64": 0.0612, "q75_var64": 0.0766, "mean_s0": 6.85, "std_s0": 3.55, "mean_s0_ratio": 1.18, "n_experts": 384 }, "shared_expert": { "var64": 0.1756, "s0": 11.97, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3382, "s0": 7.18, "s0_s1": 1.16, "shape": [ 384, 7168 ] } }, "59": { "layer": 59, "is_moe": true, "routed_experts": { "median_var64": 0.0698, "q25_var64": 0.0626, "q75_var64": 0.082, "mean_s0": 7.72, "std_s0": 5.58, "mean_s0_ratio": 1.17, "n_experts": 384 }, "shared_expert": { "var64": 0.1446, "s0": 9.76, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3457, "s0": 7.07, "s0_s1": 1.15, "shape": [ 384, 7168 ] } }, "60": { "layer": 60, "is_moe": true, "routed_experts": { "median_var64": 0.0687, "q25_var64": 0.0629, "q75_var64": 0.0769, "mean_s0": 7.74, "std_s0": 6.78, "mean_s0_ratio": 1.11, "n_experts": 384 }, "shared_expert": { "var64": 0.1227, "s0": 13.5, "shape": [ 7168, 3072 ] }, "router": { "var64": 0.3729, "s0": 8.7, "s0_s1": 1.16, "shape": [ 384, 7168 ] } } }