diff --git "a/ndarray-cache.json" "b/ndarray-cache.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache.json" @@ -0,0 +1,4395 @@ +{ + "metadata": { + "ParamSize": 327 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 65536000, + "records": [ + { + "name": "param_0", + "shape": [ + 32000, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 65536000, + "byteOffset": 0 + } + ], + "md5sum": "7684c946067246d5fff9a7fba61c6198" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_6", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "9c31dd64166c0d3d5b486ff7132de9ec" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 31784960, + "records": [ + { + "name": "param_1", + "shape": [ + 32000, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192000, + "byteOffset": 0 + }, + { + "name": "param_2", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 8192000 + }, + { + "name": "param_3", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 20774912 + }, + { + "name": "param_4", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 22347776 + }, + { + "name": "param_5", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 30736384 + } + ], + "md5sum": "2f8ce6eb075db35ae9b419b8c2398734" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_8", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "a221f7df0bcd951538b47dccdf3e6259" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_7", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_9", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_10", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_11", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_12", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_13", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "2e0b99c61d9a76ffc0b5fa76e08fec03" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_16", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "f229249e02ad9f90557c5614fbd7138e" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_18", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "7133997879279689d4a19f62efef3e75" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_14", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_15", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_17", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_19", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_20", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_21", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_22", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "0d5e5525ee1ceb34a9160464eac46d20" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_26", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "8d6e6e0697cd2b3e813ea36e2ad3ee3a" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_28", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "c602ec2b39450b3aa4cb1c5d8ccc6f25" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_23", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_24", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_25", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_27", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_29", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_30", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_31", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "d4e927d31908b43e3d6ed59a708f6f28" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_36", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "88bdfc6402e6714960e5cfd220120603" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_38", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "e99df54e84edd83a3dba3f442e997d8c" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_32", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_33", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_34", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_35", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_37", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "508d0f6919cf25b08d37fe6976996085" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_46", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "eec38d69e46f00466ce203105c751dd0" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_39", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_40", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_41", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_42", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_43", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_44", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_45", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "238ad56b418ba53a91214b76d1721364" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_48", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "fc036d31503d4df1f998987708a0940e" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_47", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_49", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_50", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_51", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_52", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_53", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "b095cbd9591ed96c1ebf9667776e0ff2" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_56", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "f015d1b3ed2e8ae82c6fee74d0c972c2" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_58", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "6fed3fc3dec8d61c2d96ffc5c01796d4" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_54", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_55", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_57", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_59", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_60", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_61", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_62", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "d72eae33997fe17127d163ab8519a63c" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_66", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "41a3eb1204fea1b6c4bed827e8ab6eac" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_68", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "57d157efbdede2722d8e6b45c1004dd3" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_63", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_64", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_65", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_67", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_69", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_70", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_71", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "c4837f9ad31f9b4386ab8fa4af4cc26d" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_76", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "40e4cf772d2d1195014b9486d982aeaa" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_78", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "3e1688e61f35f797971f8a196dc1cbf8" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_72", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_73", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_74", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_75", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_77", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "ee6558b7ea04f303ed929568c0f6b7ae" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_86", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "c3fdc4eb4574fa7579d4ae3143b03aa5" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_79", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_80", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_81", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_82", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_83", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_84", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_85", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "ea7385380c24be54e4506719b3235b16" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_88", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "1c50b0b8e33eb27997d5564ffc60d553" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_87", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_89", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_90", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_91", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_92", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_93", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "03f865c086c15c115589b759b50c385c" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_96", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "8a6aa440309201adc395ff73aadc5ff2" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_98", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "8dd9f9f5ebfac33a9fbbe4c7b79664c9" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_94", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_95", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_97", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_99", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_100", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_101", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_102", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "e201eadbb9a14952671940f18e36b6b5" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_106", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "6e80dcb377c2660b8617971328cf5b3f" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_108", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "d67d007746c52ebf882ea271cae3a5d8" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_103", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_104", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_105", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_107", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_109", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_110", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_111", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "9c80759cfbe0628139ff632997a7ddb5" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_116", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "aa95785ec9a9768e7100644170628e79" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_118", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "342509e0bfa2482b81571b0bff614131" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_112", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_113", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_114", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_115", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_117", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "c18f6a9a5e049836ad2a2baaeb248027" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_126", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "a94168c9801d62706ad0afc3cf9ba21c" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_119", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_120", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_121", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_122", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_123", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_124", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_125", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "a772b2602fafad6ec468febf8b686905" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_128", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "fc01486fd513e13cf893fd73c73003af" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_127", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_129", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_130", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_131", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_132", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_133", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "09e5ad4da322a599ef6af6ca5ae7bd54" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_136", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "95f95ad8c77795f9633517d5321acf33" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_138", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "329ee9119df8e7eacae311dd5247da54" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_134", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_135", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_137", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_139", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_140", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_141", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_142", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "1a53bfb2ed94192a3fd487b62f375a3c" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_146", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "3e9c1a1cfe6022ede2bd0da9a71ea652" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_148", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "d555e05c904e8b88d6fc22bdc05735fe" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_143", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_144", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_145", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_147", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_149", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_150", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_151", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "989961688f97a89b8a25dd2f84b2a71a" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_156", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "f2aa8267954d07f282b471053275df59" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_158", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "fc3ed1482955d878a0610809aca59bd8" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_152", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_153", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_154", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_155", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_157", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "1cc9111acfb6dc11dd8d109a5706be72" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_166", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "efdd4e86d9a81f5f9b62f7213630d0e3" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_159", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_160", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_161", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_162", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_163", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_164", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_165", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "74d9457f101e3f27a4aa1d84b023d1d1" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_168", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "3fb5b37276b0f563e96e46da8d948064" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_167", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_169", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_170", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_171", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_172", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_173", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "39a9a4946937f9fa70dfc9876e0577c4" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_176", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "ae950da37e6ebed7ef27aa5ae09ac96f" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_178", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "e01b78245a787d8027b97530bbf02ce9" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_174", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_175", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_177", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_179", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_180", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_181", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_182", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "8ba317160aaf9abb2fa78405908c0a08" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_186", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "0754e9b09ae1f248542a3144e3e5ed30" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_188", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "05f9a0928211cfc82e8a4a93e63bed07" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_183", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_184", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_185", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_187", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_189", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_190", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_191", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "22028374e0cd413b01d90fc638e85e84" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_196", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "fc38283d6bf3c4c46bbd1059ff65fa34" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_198", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "6c0b0bb04f701a81d7325768fefb3567" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_192", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_193", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_194", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_195", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_197", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "1431e552e06729cf2afb1e5267bdd9d7" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_206", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "6acc26b161143e30140fc2f30c265a0e" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_199", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_200", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_201", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_202", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_203", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_204", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_205", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "c790af2200a8a0ae244ebbce7b3e2e98" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_208", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "d1d096d823b98e0686506d5acf3099f3" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_207", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_209", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_210", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_211", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_212", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_213", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "69dfee4799ae7854983669670c5812ca" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_216", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "d46ed7e1f8e43dd2063effd9c72e39b1" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_218", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "5d863ded3037b767b9b3dcf4c16aa7db" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_214", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_215", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_217", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_219", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_220", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_221", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_222", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "e6cbefa8004d5168981cf087c84ee791" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_226", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "4e5ee41bea20a131c4b6149a4a98dfb6" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_228", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "b0ab4dc068d1ce6b05ac4317bf10aa80" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_223", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_224", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_225", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_227", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_229", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_230", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_231", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "c0b2c8b0aeb09fee8c88f0c09de1e245" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_236", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "32c215e62a9e256893c6a7e237efdf41" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_238", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "1ff018016032644c09db6eda76e0fbc9" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_232", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_233", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_234", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_235", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_237", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "711a8dd91884d87834a7d48d56b0ad3e" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_246", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "b64f8dc9d31f020183f3b5b52d7de54c" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_239", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_240", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_241", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_242", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_243", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_244", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_245", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "e419d89ba013f0556d0c5c28b0366655" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_248", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "e6952eda770d2949f97368c1a46b8ef6" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_247", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_249", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_250", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_251", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_252", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_253", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "f21ab9a907373f42a76d159767e687a1" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_256", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "82e8df9aecf5313b94aff6d58bcbecb1" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_258", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "f729b0394a96e1c1600c135a8bc2f377" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_254", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_255", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_257", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_259", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_260", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_261", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_262", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "102ca24be6316bb314e68dc402bae60c" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_266", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "b54a90e1cc8c702970faa4fca018b1ce" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_268", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "ffb166a03d5dedd621ec66172445c95d" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_263", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_264", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_265", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_267", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_269", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_270", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_271", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "4f71f455ba183e91b1515ad8996de0e6" + }, + { + "dataPath": "params_shard_89.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_276", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "b47d096a4d53f13fe78c86b716b3aa98" + }, + { + "dataPath": "params_shard_90.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_278", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "7a47909cfc2e3524d9f92f95365a38d1" + }, + { + "dataPath": "params_shard_91.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_272", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_273", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_274", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_275", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_277", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "3b147fd28722f4d8724dddedc3bce89b" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_286", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "958b36d6a92c7668efefb7030ff038de" + }, + { + "dataPath": "params_shard_93.bin", + "format": "raw-shard", + "nbytes": 27279360, + "records": [ + { + "name": "param_279", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_280", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_281", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_282", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 3686400 + }, + { + "name": "param_283", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 16269312 + }, + { + "name": "param_284", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 17842176 + }, + { + "name": "param_285", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 26230784 + } + ], + "md5sum": "4ba9c2763486fab37716e90209db1997" + }, + { + "dataPath": "params_shard_94.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_288", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "bfdbb16fbc20561a5ea932cdee3eaef0" + }, + { + "dataPath": "params_shard_95.bin", + "format": "raw-shard", + "nbytes": 25182208, + "records": [ + { + "name": "param_287", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "param_289", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 7340032 + }, + { + "name": "param_290", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11010048 + }, + { + "name": "param_291", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 11018240 + }, + { + "name": "param_292", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 11026432 + }, + { + "name": "param_293", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 23609344 + } + ], + "md5sum": "158d0fe3d31b94f4ac663005de02fad3" + }, + { + "dataPath": "params_shard_96.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_296", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "b4b0219a277a3bd095f30f31f9d81620" + }, + { + "dataPath": "params_shard_97.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_298", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "982093d42bc9e2b63aa02abebda9f7d6" + }, + { + "dataPath": "params_shard_98.bin", + "format": "raw-shard", + "nbytes": 33046528, + "records": [ + { + "name": "param_294", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 0 + }, + { + "name": "param_295", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 8388608 + }, + { + "name": "param_297", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 9437184 + }, + { + "name": "param_299", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 16777216 + }, + { + "name": "param_300", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20447232 + }, + { + "name": "param_301", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 20455424 + }, + { + "name": "param_302", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 20463616 + } + ], + "md5sum": "1c86144404323bdd55b7dd6e2073deb2" + }, + { + "dataPath": "params_shard_99.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_306", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "c1b8c73fe86a97b9ae9c991ee106a6c1" + }, + { + "dataPath": "params_shard_100.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_308", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "4b369ed7ccbfc091c36e11c907758116" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 22036480, + "records": [ + { + "name": "param_303", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 0 + }, + { + "name": "param_304", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 1572864 + }, + { + "name": "param_305", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 9961472 + }, + { + "name": "param_307", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 11010048 + }, + { + "name": "param_309", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 18350080 + }, + { + "name": "param_310", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22020096 + }, + { + "name": "param_311", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 22028288 + } + ], + "md5sum": "1e8176625bb8c1188bfccefaca879635" + }, + { + "dataPath": "params_shard_102.bin", + "format": "raw-shard", + "nbytes": 58720256, + "records": [ + { + "name": "param_316", + "shape": [ + 28672, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 58720256, + "byteOffset": 0 + } + ], + "md5sum": "1b684e7b1f87f39e461bdf0d49253fae" + }, + { + "dataPath": "params_shard_103.bin", + "format": "raw-shard", + "nbytes": 29360128, + "records": [ + { + "name": "param_318", + "shape": [ + 4096, + 1792 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 29360128, + "byteOffset": 0 + } + ], + "md5sum": "6d0702052e8be9043c6f37c27e2824c2" + }, + { + "dataPath": "params_shard_104.bin", + "format": "raw-shard", + "nbytes": 30932992, + "records": [ + { + "name": "param_312", + "shape": [ + 6144, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "param_313", + "shape": [ + 6144, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "param_314", + "shape": [ + 4096, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 8388608, + "byteOffset": 14155776 + }, + { + "name": "param_315", + "shape": [ + 4096, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 1048576, + "byteOffset": 22544384 + }, + { + "name": "param_317", + "shape": [ + 28672, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 7340032, + "byteOffset": 23592960 + } + ], + "md5sum": "fc49d1f696c56c295c6f7fa53af593c0" + }, + { + "dataPath": "params_shard_105.bin", + "format": "raw-shard", + "nbytes": 65536000, + "records": [ + { + "name": "param_323", + "shape": [ + 32000, + 512 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 65536000, + "byteOffset": 0 + } + ], + "md5sum": "2a60b11f53fde2380ea535df319c423a" + }, + { + "dataPath": "params_shard_106.bin", + "format": "raw-shard", + "nbytes": 12935168, + "records": [ + { + "name": "param_319", + "shape": [ + 4096, + 448 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 3670016, + "byteOffset": 0 + }, + { + "name": "param_320", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3670016 + }, + { + "name": "param_321", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3678208 + }, + { + "name": "param_322", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192, + "byteOffset": 3686400 + }, + { + "name": "param_324", + "shape": [ + 32000, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 8192000, + "byteOffset": 3694592 + }, + { + "name": "param_325", + "shape": [ + 2048, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 524288, + "byteOffset": 11886592 + }, + { + "name": "param_326", + "shape": [ + 2048, + 128 + ], + "dtype": "float16", + "format": "raw", + "nbytes": 524288, + "byteOffset": 12410880 + } + ], + "md5sum": "8e3ff3dedf9fc0be81b3466618188616" + } + ] +} \ No newline at end of file