{ "version": 2, "model_id": "moonshotai/Kimi-K2-Instruct", "model_config": { "moe": { "model_type": "kimi_k2", "num_layers": 61, "hidden_size": 7168, "intermediate_size": 18432, "moe_intermediate_size": 2048, "n_routed_experts": 384, "n_shared_experts": 1, "num_experts_per_tok": 8, "first_k_dense_replace": 1, "torch_dtype": "bfloat16", "quant_method": "fp8" } }, "num_feats": 64, "dtype": "float32", "gate_vector_semantics": "right_singular_vectors_of_gate_proj", "layers": { "1": { "shape": [ 384, 64, 7168 ], "file_offset": 0 }, "2": { "shape": [ 384, 64, 7168 ], "file_offset": 704643072 }, "3": { "shape": [ 384, 64, 7168 ], "file_offset": 1409286144 }, "4": { "shape": [ 384, 64, 7168 ], "file_offset": 2113929216 }, "5": { "shape": [ 384, 64, 7168 ], "file_offset": 2818572288 }, "6": { "shape": [ 384, 64, 7168 ], "file_offset": 3523215360 }, "7": { "shape": [ 384, 64, 7168 ], "file_offset": 4227858432 }, "8": { "shape": [ 384, 64, 7168 ], "file_offset": 4932501504 }, "9": { "shape": [ 384, 64, 7168 ], "file_offset": 5637144576 }, "10": { "shape": [ 384, 64, 7168 ], "file_offset": 6341787648 }, "11": { "shape": [ 384, 64, 7168 ], "file_offset": 7046430720 }, "12": { "shape": [ 384, 64, 7168 ], "file_offset": 7751073792 }, "13": { "shape": [ 384, 64, 7168 ], "file_offset": 8455716864 }, "14": { "shape": [ 384, 64, 7168 ], "file_offset": 9160359936 }, "15": { "shape": [ 384, 64, 7168 ], "file_offset": 9865003008 }, "16": { "shape": [ 384, 64, 7168 ], "file_offset": 10569646080 }, "17": { "shape": [ 384, 64, 7168 ], "file_offset": 11274289152 }, "18": { "shape": [ 384, 64, 7168 ], "file_offset": 11978932224 }, "19": { "shape": [ 384, 64, 7168 ], "file_offset": 12683575296 }, "20": { "shape": [ 384, 64, 7168 ], "file_offset": 13388218368 }, "21": { "shape": [ 384, 64, 7168 ], "file_offset": 14092861440 }, "22": { "shape": [ 384, 64, 7168 ], "file_offset": 14797504512 }, "23": { "shape": [ 384, 64, 7168 ], "file_offset": 15502147584 }, "24": { "shape": [ 384, 64, 7168 ], "file_offset": 16206790656 }, "25": { "shape": [ 384, 64, 7168 ], "file_offset": 16911433728 }, "26": { "shape": [ 384, 64, 7168 ], "file_offset": 17616076800 }, "27": { "shape": [ 384, 64, 7168 ], "file_offset": 18320719872 }, "28": { "shape": [ 384, 64, 7168 ], "file_offset": 19025362944 }, "29": { "shape": [ 384, 64, 7168 ], "file_offset": 19730006016 }, "30": { "shape": [ 384, 64, 7168 ], "file_offset": 20434649088 }, "31": { "shape": [ 384, 64, 7168 ], "file_offset": 21139292160 }, "32": { "shape": [ 384, 64, 7168 ], "file_offset": 21843935232 }, "33": { "shape": [ 384, 64, 7168 ], "file_offset": 22548578304 }, "34": { "shape": [ 384, 64, 7168 ], "file_offset": 23253221376 }, "35": { "shape": [ 384, 64, 7168 ], "file_offset": 23957864448 }, "36": { "shape": [ 384, 64, 7168 ], "file_offset": 24662507520 }, "37": { "shape": [ 384, 64, 7168 ], "file_offset": 25367150592 }, "38": { "shape": [ 384, 64, 7168 ], "file_offset": 26071793664 }, "39": { "shape": [ 384, 64, 7168 ], "file_offset": 26776436736 }, "40": { "shape": [ 384, 64, 7168 ], "file_offset": 27481079808 }, "41": { "shape": [ 384, 64, 7168 ], "file_offset": 28185722880 }, "42": { "shape": [ 384, 64, 7168 ], "file_offset": 28890365952 }, "43": { "shape": [ 384, 64, 7168 ], "file_offset": 29595009024 }, "44": { "shape": [ 384, 64, 7168 ], "file_offset": 30299652096 }, "45": { "shape": [ 384, 64, 7168 ], "file_offset": 31004295168 }, "46": { "shape": [ 384, 64, 7168 ], "file_offset": 31708938240 }, "47": { "shape": [ 384, 64, 7168 ], "file_offset": 32413581312 }, "48": { "shape": [ 384, 64, 7168 ], "file_offset": 33118224384 }, "49": { "shape": [ 384, 64, 7168 ], "file_offset": 33822867456 }, "50": { "shape": [ 384, 64, 7168 ], "file_offset": 34527510528 }, "51": { "shape": [ 384, 64, 7168 ], "file_offset": 35232153600 }, "52": { "shape": [ 384, 64, 7168 ], "file_offset": 35936796672 }, "53": { "shape": [ 384, 64, 7168 ], "file_offset": 36641439744 }, "54": { "shape": [ 384, 64, 7168 ], "file_offset": 37346082816 }, "55": { "shape": [ 384, 64, 7168 ], "file_offset": 38050725888 }, "56": { "shape": [ 384, 64, 7168 ], "file_offset": 38755368960 }, "57": { "shape": [ 384, 64, 7168 ], "file_offset": 39460012032 }, "58": { "shape": [ 384, 64, 7168 ], "file_offset": 40164655104 }, "59": { "shape": [ 384, 64, 7168 ], "file_offset": 40869298176 }, "60": { "shape": [ 384, 64, 7168 ], "file_offset": 41573941248 } }, "layer_stats": { "1": { "median_var64": 0.0946, "q25_var64": 0.0858, "q75_var64": 0.1024, "n_experts": 384 }, "2": { "median_var64": 0.0968, "q25_var64": 0.0927, "q75_var64": 0.1005, "n_experts": 384 }, "3": { "median_var64": 0.0954, "q25_var64": 0.0927, "q75_var64": 0.0991, "n_experts": 384 }, "4": { "median_var64": 0.0963, "q25_var64": 0.0928, "q75_var64": 0.1008, "n_experts": 384 }, "5": { "median_var64": 0.098, "q25_var64": 0.0944, "q75_var64": 0.1022, "n_experts": 384 }, "6": { "median_var64": 0.0944, "q25_var64": 0.0913, "q75_var64": 0.0975, "n_experts": 384 }, "7": { "median_var64": 0.0925, "q25_var64": 0.0895, "q75_var64": 0.0965, "n_experts": 384 }, "8": { "median_var64": 0.0925, "q25_var64": 0.0889, "q75_var64": 0.0971, "n_experts": 384 }, "9": { "median_var64": 0.091, "q25_var64": 0.0869, "q75_var64": 0.096, "n_experts": 384 }, "10": { "median_var64": 0.0926, "q25_var64": 0.0876, "q75_var64": 0.0984, "n_experts": 384 }, "11": { "median_var64": 0.0927, "q25_var64": 0.0874, "q75_var64": 0.0996, "n_experts": 384 }, "12": { "median_var64": 0.0961, "q25_var64": 0.089, "q75_var64": 0.1033, "n_experts": 384 }, "13": { "median_var64": 0.1009, "q25_var64": 0.0925, "q75_var64": 0.1102, "n_experts": 384 }, "14": { "median_var64": 0.1015, "q25_var64": 0.0919, "q75_var64": 0.1102, "n_experts": 384 }, "15": { "median_var64": 0.1003, "q25_var64": 0.0913, "q75_var64": 0.1137, "n_experts": 384 }, "16": { "median_var64": 0.1007, "q25_var64": 0.0912, "q75_var64": 0.1146, "n_experts": 384 }, "17": { "median_var64": 0.1047, "q25_var64": 0.0924, "q75_var64": 0.1171, "n_experts": 384 }, "18": { "median_var64": 0.1034, "q25_var64": 0.0902, "q75_var64": 0.1158, "n_experts": 384 }, "19": { "median_var64": 0.1018, "q25_var64": 0.0888, "q75_var64": 0.1168, "n_experts": 384 }, "20": { "median_var64": 0.1023, "q25_var64": 0.0882, "q75_var64": 0.1181, "n_experts": 384 }, "21": { "median_var64": 0.1061, "q25_var64": 0.09, "q75_var64": 0.1209, "n_experts": 384 }, "22": { "median_var64": 0.106, "q25_var64": 0.09, "q75_var64": 0.1248, "n_experts": 384 }, "23": { "median_var64": 0.1065, "q25_var64": 0.0882, "q75_var64": 0.1258, "n_experts": 384 }, "24": { "median_var64": 0.1079, "q25_var64": 0.0916, "q75_var64": 0.1298, "n_experts": 384 }, "25": { "median_var64": 0.106, "q25_var64": 0.0895, "q75_var64": 0.1251, "n_experts": 384 }, "26": { "median_var64": 0.1042, "q25_var64": 0.09, "q75_var64": 0.1263, "n_experts": 384 }, "27": { "median_var64": 0.1013, "q25_var64": 0.089, "q75_var64": 0.1229, "n_experts": 384 }, "28": { "median_var64": 0.0995, "q25_var64": 0.0875, "q75_var64": 0.1186, "n_experts": 384 }, "29": { "median_var64": 0.1005, "q25_var64": 0.0873, "q75_var64": 0.1157, "n_experts": 384 }, "30": { "median_var64": 0.0955, "q25_var64": 0.0864, "q75_var64": 0.1129, "n_experts": 384 }, "31": { "median_var64": 0.0927, "q25_var64": 0.0856, "q75_var64": 0.1083, "n_experts": 384 }, "32": { "median_var64": 0.09, "q25_var64": 0.0843, "q75_var64": 0.1049, "n_experts": 384 }, "33": { "median_var64": 0.0903, "q25_var64": 0.0833, "q75_var64": 0.1028, "n_experts": 384 }, "34": { "median_var64": 0.0893, "q25_var64": 0.0844, "q75_var64": 0.0996, "n_experts": 384 }, "35": { "median_var64": 0.0874, "q25_var64": 0.0829, "q75_var64": 0.0984, "n_experts": 384 }, "36": { "median_var64": 0.0879, "q25_var64": 0.083, "q75_var64": 0.0988, "n_experts": 384 }, "37": { "median_var64": 0.0866, "q25_var64": 0.0821, "q75_var64": 0.0966, "n_experts": 384 }, "38": { "median_var64": 0.0867, "q25_var64": 0.0824, "q75_var64": 0.0939, "n_experts": 384 }, "39": { "median_var64": 0.0848, "q25_var64": 0.0812, "q75_var64": 0.0916, "n_experts": 384 }, "40": { "median_var64": 0.0835, "q25_var64": 0.0806, "q75_var64": 0.0903, "n_experts": 384 }, "41": { "median_var64": 0.0832, "q25_var64": 0.0802, "q75_var64": 0.0892, "n_experts": 384 }, "42": { "median_var64": 0.0832, "q25_var64": 0.0804, "q75_var64": 0.0898, "n_experts": 384 }, "43": { "median_var64": 0.083, "q25_var64": 0.0804, "q75_var64": 0.0892, "n_experts": 384 }, "44": { "median_var64": 0.0832, "q25_var64": 0.0805, "q75_var64": 0.0905, "n_experts": 384 }, "45": { "median_var64": 0.0831, "q25_var64": 0.0802, "q75_var64": 0.0889, "n_experts": 384 }, "46": { "median_var64": 0.0834, "q25_var64": 0.0802, "q75_var64": 0.0907, "n_experts": 384 }, "47": { "median_var64": 0.0833, "q25_var64": 0.0799, "q75_var64": 0.0901, "n_experts": 384 }, "48": { "median_var64": 0.0834, "q25_var64": 0.0798, "q75_var64": 0.0901, "n_experts": 384 }, "49": { "median_var64": 0.0835, "q25_var64": 0.0799, "q75_var64": 0.0921, "n_experts": 384 }, "50": { "median_var64": 0.0848, "q25_var64": 0.0806, "q75_var64": 0.0921, "n_experts": 384 }, "51": { "median_var64": 0.0856, "q25_var64": 0.0808, "q75_var64": 0.0951, "n_experts": 384 }, "52": { "median_var64": 0.0872, "q25_var64": 0.081, "q75_var64": 0.097, "n_experts": 384 }, "53": { "median_var64": 0.0872, "q25_var64": 0.0817, "q75_var64": 0.1009, "n_experts": 384 }, "54": { "median_var64": 0.0911, "q25_var64": 0.0827, "q75_var64": 0.1038, "n_experts": 384 }, "55": { "median_var64": 0.0932, "q25_var64": 0.0845, "q75_var64": 0.1058, "n_experts": 384 }, "56": { "median_var64": 0.0944, "q25_var64": 0.0844, "q75_var64": 0.109, "n_experts": 384 }, "57": { "median_var64": 0.096, "q25_var64": 0.0873, "q75_var64": 0.1137, "n_experts": 384 }, "58": { "median_var64": 0.0959, "q25_var64": 0.0863, "q75_var64": 0.1112, "n_experts": 384 }, "59": { "median_var64": 0.0955, "q25_var64": 0.0866, "q75_var64": 0.1091, "n_experts": 384 }, "60": { "median_var64": 0.0975, "q25_var64": 0.0884, "q75_var64": 0.1075, "n_experts": 384 } } }