{ "metadata": { "ParamSize": 455, "ParamBytes": 1252341760.0, "BitsPerParam": 3.6042710814191015 }, "records": [ { "dataPath": "params_shard_0.bin", "format": "raw-shard", "nbytes": 52428800, "records": [ { "name": "lm_head.linear.q_weight", "shape": [ 51200, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 52428800, "byteOffset": 0 } ], "md5sum": "8503cdb31752d6d04435cc790e885ef8" }, { "dataPath": "params_shard_1.bin", "format": "raw-shard", "nbytes": 30305280, "records": [ { "name": "lm_head.linear.bias", "shape": [ 51200 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 102400, "byteOffset": 0 }, { "name": "lm_head.linear.q_scale", "shape": [ 51200, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6553600, "byteOffset": 102400 }, { "name": "lm_head.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 6656000 }, { "name": "lm_head.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 6661120 }, { "name": "transformer.h.30.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 6666240 }, { "name": "transformer.h.30.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 6681600 }, { "name": "transformer.h.30.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 14545920 }, { "name": "transformer.h.30.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 15528960 }, { "name": "transformer.h.30.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 15534080 }, { "name": "transformer.h.30.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 18155520 }, { "name": "transformer.h.30.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 18483200 }, { "name": "transformer.h.30.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 18503680 }, { "name": "transformer.h.30.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 28989440 }, { "name": "transformer.h.30.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 30300160 } ], "md5sum": "51daf22d278290a6ef9e2cb4c372c731" }, { "dataPath": "params_shard_2.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.30.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.30.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.31.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.31.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.31.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.31.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.31.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.31.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.31.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.31.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.31.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "220009150531dd089a3ff5c991439f8b" }, { "dataPath": "params_shard_3.bin", "format": "raw-shard", "nbytes": 52428800, "records": [ { "name": "transformer.embd.q_weight", "shape": [ 51200, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 52428800, "byteOffset": 0 } ], "md5sum": "444a1ce65cfb68d42136ebe38da9381a" }, { "dataPath": "params_shard_4.bin", "format": "raw-shard", "nbytes": 30177280, "records": [ { "name": "transformer.h.31.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.31.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.31.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.31.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.31.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.embd.q_scale", "shape": [ 51200, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6553600, "byteOffset": 23598080 }, { "name": "transformer.h.0.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 30151680 }, { "name": "transformer.h.0.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 30156800 }, { "name": "transformer.h.0.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 30161920 } ], "md5sum": "3bb78f1b1e593bd98d4217cedf19191c" }, { "dataPath": "params_shard_5.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.0.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.0.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.0.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.0.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.0.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.0.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.0.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.0.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.0.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "820c7648fe62987900bcd969016cf1bb" }, { "dataPath": "params_shard_6.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.0.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.0.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.1.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.1.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.1.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.1.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.1.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.1.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.1.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.1.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.1.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "1bf257dfcb0dacaa41167ae4feeb5b45" }, { "dataPath": "params_shard_7.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.1.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.1.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.1.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.1.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.1.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.10.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.10.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.10.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.10.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.10.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.10.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "51a6f838c8734a64a23ef66a2d02fdde" }, { "dataPath": "params_shard_8.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.10.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.10.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.10.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.10.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.10.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.10.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.10.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.10.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.11.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.11.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.11.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "8b310759583a154460453b4ef06404f2" }, { "dataPath": "params_shard_9.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.11.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.11.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.11.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.11.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.11.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.11.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.11.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.11.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.11.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "89552acc24a7847330313349a139093e" }, { "dataPath": "params_shard_10.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.11.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.11.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.12.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.12.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.12.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.12.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.12.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.12.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.12.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.12.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.12.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "0a41aa89ea18701784e6ac93c7a83ab5" }, { "dataPath": "params_shard_11.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.12.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.12.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.12.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.12.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.12.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.13.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.13.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.13.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.13.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.13.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.13.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "43faab711830b9ea49e3194e846bf628" }, { "dataPath": "params_shard_12.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.13.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.13.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.13.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.13.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.13.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.13.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.13.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.13.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.14.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.14.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.14.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "80bf75db9e545dc2169d2ccf624da9ef" }, { "dataPath": "params_shard_13.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.14.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.14.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.14.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.14.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.14.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.14.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.14.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.14.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.14.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "bbaf0debc8f982e8fcd67490b925ef81" }, { "dataPath": "params_shard_14.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.14.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.14.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.15.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.15.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.15.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.15.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.15.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.15.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.15.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.15.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.15.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "cde2fc0462e3f9b39828dc11295e465f" }, { "dataPath": "params_shard_15.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.15.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.15.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.15.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.15.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.15.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.16.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.16.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.16.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.16.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.16.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.16.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "9aa5ca0761e73307f100a685632bbeea" }, { "dataPath": "params_shard_16.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.16.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.16.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.16.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.16.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.16.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.16.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.16.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.16.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.17.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.17.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.17.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "bc821dfd4b18d47e408cc50ab55d988b" }, { "dataPath": "params_shard_17.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.17.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.17.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.17.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.17.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.17.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.17.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.17.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.17.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.17.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "37995a6caca01601cd7efa8cacecf833" }, { "dataPath": "params_shard_18.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.17.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.17.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.18.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.18.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.18.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.18.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.18.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.18.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.18.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.18.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.18.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "2dfb8cf8657c5e804e79e902beb8874e" }, { "dataPath": "params_shard_19.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.18.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.18.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.18.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.18.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.18.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.19.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.19.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.19.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.19.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.19.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.19.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "e1828b0dfd194a1304eb64b2a3d6ace7" }, { "dataPath": "params_shard_20.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.19.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.19.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.19.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.19.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.19.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.19.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.19.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.19.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.2.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.2.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.2.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "756a0bf111183079ff599585e64ec58d" }, { "dataPath": "params_shard_21.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.2.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.2.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.2.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.2.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.2.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.2.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.2.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.2.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.2.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "4d3f6284ab57c7a2b3a2e288e5191019" }, { "dataPath": "params_shard_22.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.2.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.2.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.20.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.20.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.20.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.20.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.20.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.20.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.20.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.20.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.20.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "db2eab185667fca5e71f35a8193cec85" }, { "dataPath": "params_shard_23.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.20.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.20.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.20.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.20.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.20.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.21.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.21.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.21.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.21.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.21.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.21.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "f7869a670740d5cf77ae277756a046c9" }, { "dataPath": "params_shard_24.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.21.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.21.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.21.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.21.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.21.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.21.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.21.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.21.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.22.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.22.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.22.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "ae10f14690d9d0c4af86869a6de6e639" }, { "dataPath": "params_shard_25.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.22.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.22.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.22.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.22.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.22.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.22.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.22.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.22.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.22.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "621b33076777a15eab91006f7fcf9632" }, { "dataPath": "params_shard_26.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.22.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.22.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.23.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.23.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.23.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.23.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.23.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.23.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.23.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.23.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.23.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "0f56328d078938f242d0ba888614d204" }, { "dataPath": "params_shard_27.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.23.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.23.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.23.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.23.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.23.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.24.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.24.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.24.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.24.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.24.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.24.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "b42a9fc2fd28fc85c7cc1a168dab6c77" }, { "dataPath": "params_shard_28.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.24.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.24.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.24.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.24.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.24.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.24.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.24.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.24.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.25.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.25.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.25.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "b1d0dd47a2c2279c4a5b15a8f8bbbe34" }, { "dataPath": "params_shard_29.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.25.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.25.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.25.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.25.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.25.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.25.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.25.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.25.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.25.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "052d166da83b518f0e07b2ca99e11ede" }, { "dataPath": "params_shard_30.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.25.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.25.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.26.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.26.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.26.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.26.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.26.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.26.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.26.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.26.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.26.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "c66734017f96495525e5eddfd68e1e32" }, { "dataPath": "params_shard_31.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.26.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.26.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.26.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.26.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.26.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.27.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.27.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.27.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.27.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.27.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.27.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "7d903cb9a923785d82c1a2e554482782" }, { "dataPath": "params_shard_32.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.27.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.27.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.27.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.27.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.27.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.27.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.27.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.27.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.28.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.28.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.28.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "00491905181e6300d3d6246f91054b51" }, { "dataPath": "params_shard_33.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.28.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.28.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.28.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.28.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.28.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.28.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.28.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.28.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.28.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "60b2dd4314c97fd4662491932d5fbb74" }, { "dataPath": "params_shard_34.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.28.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.28.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.29.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.29.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.29.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.29.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.29.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.29.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.29.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.29.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.29.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "3648d31c36900309d4465fa22f092467" }, { "dataPath": "params_shard_35.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.29.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.29.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.29.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.29.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.29.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.3.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.3.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.3.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.3.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.3.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.3.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "b4d733d59ac113c63b3feabdd09bf4d1" }, { "dataPath": "params_shard_36.bin", "format": "raw-shard", "nbytes": 26603520, "records": [ { "name": "transformer.h.3.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.3.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.3.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.3.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.3.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.3.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.3.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.3.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.30.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.30.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.4.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26577920 }, { "name": "transformer.h.4.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26583040 }, { "name": "transformer.h.4.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26588160 } ], "md5sum": "e7737b15ba5f7da730a2bb6cab8c66dc" }, { "dataPath": "params_shard_37.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.4.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.4.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.4.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.4.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.4.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.4.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.4.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.4.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.4.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "0597d18af46a583ff35fc0e0a4b0ae06" }, { "dataPath": "params_shard_38.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.4.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.4.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.5.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.5.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.5.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.5.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.5.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.5.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.5.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.5.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.5.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "77adda8104cb0b0582896e5344a655e2" }, { "dataPath": "params_shard_39.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.5.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.5.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.5.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.5.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.5.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.6.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.6.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.6.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.6.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.6.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.6.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "8ac89b1b02094e0a34e890e478e520e6" }, { "dataPath": "params_shard_40.bin", "format": "raw-shard", "nbytes": 26593280, "records": [ { "name": "transformer.h.6.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.6.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.6.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.6.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.6.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.6.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.6.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.6.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 }, { "name": "transformer.h.7.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26567680 }, { "name": "transformer.h.7.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 26572800 }, { "name": "transformer.h.7.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 26577920 } ], "md5sum": "241b692a71cb3d5ada05a33c5b1e8717" }, { "dataPath": "params_shard_41.bin", "format": "raw-shard", "nbytes": 23623680, "records": [ { "name": "transformer.h.7.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 0 }, { "name": "transformer.h.7.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 7864320 }, { "name": "transformer.h.7.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8847360 }, { "name": "transformer.h.7.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 8852480 }, { "name": "transformer.h.7.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 11473920 }, { "name": "transformer.h.7.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 11801600 }, { "name": "transformer.h.7.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11822080 }, { "name": "transformer.h.7.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22307840 }, { "name": "transformer.h.7.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23618560 } ], "md5sum": "28de01ca76153d1044460eb214ef2e76" }, { "dataPath": "params_shard_42.bin", "format": "raw-shard", "nbytes": 23644160, "records": [ { "name": "transformer.h.7.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.7.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.8.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.8.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11801600 }, { "name": "transformer.h.8.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 11806720 }, { "name": "transformer.h.8.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 11822080 }, { "name": "transformer.h.8.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 19686400 }, { "name": "transformer.h.8.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 20669440 }, { "name": "transformer.h.8.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 20674560 }, { "name": "transformer.h.8.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 23296000 }, { "name": "transformer.h.8.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23623680 } ], "md5sum": "9a90bb2f472a31878b452fa00b045506" }, { "dataPath": "params_shard_43.bin", "format": "raw-shard", "nbytes": 32476160, "records": [ { "name": "transformer.h.8.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 0 }, { "name": "transformer.h.8.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 10485760 }, { "name": "transformer.h.8.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11796480 }, { "name": "transformer.h.8.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 11801600 }, { "name": "transformer.h.8.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 22287360 }, { "name": "transformer.h.9.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23598080 }, { "name": "transformer.h.9.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 23603200 }, { "name": "transformer.h.9.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 23608320 }, { "name": "transformer.h.9.mixer.Wqkv.q_weight", "shape": [ 7680, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 7864320, "byteOffset": 23623680 }, { "name": "transformer.h.9.mixer.Wqkv.q_scale", "shape": [ 7680, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 983040, "byteOffset": 31488000 }, { "name": "transformer.h.9.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 32471040 } ], "md5sum": "8f8fba896332a98cba11fd5581b8c681" }, { "dataPath": "params_shard_44.bin", "format": "raw-shard", "nbytes": 26567680, "records": [ { "name": "transformer.h.9.mixer.out_proj.q_weight", "shape": [ 2560, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 2621440, "byteOffset": 0 }, { "name": "transformer.h.9.mixer.out_proj.q_scale", "shape": [ 2560, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 327680, "byteOffset": 2621440 }, { "name": "transformer.h.9.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 2949120 }, { "name": "transformer.h.9.mlp.fc1.q_weight", "shape": [ 10240, 256 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 2969600 }, { "name": "transformer.h.9.mlp.fc1.q_scale", "shape": [ 10240, 64 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 13455360 }, { "name": "transformer.h.9.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14766080 }, { "name": "transformer.h.9.mlp.fc2.q_weight", "shape": [ 2560, 1024 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 10485760, "byteOffset": 14771200 }, { "name": "transformer.h.9.mlp.fc2.q_scale", "shape": [ 2560, 256 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1310720, "byteOffset": 25256960 } ], "md5sum": "3e99440d37f01206f3f1272cb82b84aa" } ] }