{ "metadata": { "ParamSize": 325, "ParamBytes": 2149644288.0, "BitsPerParam": 4.500600961055312 }, "records": [ { "dataPath": "params_shard_0.bin", "format": "raw-shard", "nbytes": 49250304, "records": [ { "name": "lm_head.q_weight", "shape": [ 384, 32064 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 49250304, "byteOffset": 0 } ], "md5sum": "51778898914f5112659c6783e6c8c812" }, { "dataPath": "params_shard_1.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.21.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "046882c84cd2cbd5f05bedab16db7cf7" }, { "dataPath": "params_shard_2.bin", "format": "raw-shard", "nbytes": 23470080, "records": [ { "name": "lm_head.q_scale", "shape": [ 96, 32064 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6156288, "byteOffset": 0 }, { "name": "transformer.h.21.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 6156288 }, { "name": "transformer.h.21.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 6162432 }, { "name": "transformer.h.21.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 18745344 }, { "name": "transformer.h.21.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 20318208 }, { "name": "transformer.h.21.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 23463936 } ], "md5sum": "a318ee36fe27a3e3408f23f2a85d3f81" }, { "dataPath": "params_shard_3.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.22.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "ce4b1fe36fd1f02610cbd1e78cce7054" }, { "dataPath": "params_shard_4.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.21.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.21.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.22.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.22.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.22.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.22.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.22.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "8d57c5dd32477d9c4aa96e2a0cfcc05b" }, { "dataPath": "params_shard_5.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.22.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.22.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.22.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.22.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.23.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "e258e9077c1dc3a83eda7c411602b0a5" }, { "dataPath": "params_shard_6.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.23.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "ffb49c03cdc14afd5a5cbc03c94837d7" }, { "dataPath": "params_shard_7.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.23.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.23.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.23.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.23.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.23.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.23.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "414b283d64920106e745cde49bfa4888" }, { "dataPath": "params_shard_8.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.24.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "9758958ab1423d51b9afa05f224524f2" }, { "dataPath": "params_shard_9.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.23.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.23.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.24.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.24.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.24.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.24.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.24.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "c677e252462373f7f41b74f9bb01a525" }, { "dataPath": "params_shard_10.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.24.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.24.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.24.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.24.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.25.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "3d81a487048dbd3c8e5240dd1fa833f8" }, { "dataPath": "params_shard_11.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.25.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "7426646b13fe843d89b4675a8114603d" }, { "dataPath": "params_shard_12.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.25.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.25.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.25.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.25.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.25.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.25.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "906a933e661463134902d323ec902f9a" }, { "dataPath": "params_shard_13.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.26.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "19fb8b42431b0942110119058570dcad" }, { "dataPath": "params_shard_14.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.25.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.25.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.26.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.26.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.26.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.26.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.26.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "7e2faa5a4d9369723b0ac846056ab4c3" }, { "dataPath": "params_shard_15.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.26.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.26.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.26.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.26.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.27.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "9e3003b86c1c50bc148e4881d08ccf60" }, { "dataPath": "params_shard_16.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.27.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "02a9f7036f086e6d988dc716cffae342" }, { "dataPath": "params_shard_17.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.27.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.27.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.27.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.27.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.27.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.27.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "aa2cece5bbea0472eb2d4859362b98fa" }, { "dataPath": "params_shard_18.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.28.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "311f752ca200e3eaa6d7477243237a69" }, { "dataPath": "params_shard_19.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.27.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.27.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.28.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.28.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.28.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.28.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.28.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "49e93c4bdd457cf35069e8456c88b84b" }, { "dataPath": "params_shard_20.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.28.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.28.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.28.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.28.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.29.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "235d00dd943131382d5b644b87f6b371" }, { "dataPath": "params_shard_21.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.29.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "8989e095361e8b3da2e4f84589b1f941" }, { "dataPath": "params_shard_22.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.29.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.29.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.29.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.29.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.29.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.29.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "c8b987883f46e67d23b1899d4e23bd16" }, { "dataPath": "params_shard_23.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.30.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "01c378ef6c03860ebac48f126144f8a8" }, { "dataPath": "params_shard_24.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.29.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.29.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.30.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.30.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.30.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.30.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.30.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "62b70fc1520afc5193043653ca419822" }, { "dataPath": "params_shard_25.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.30.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.30.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.30.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.30.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.31.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "76222a5132ba8156eecb93bbf8ba40ff" }, { "dataPath": "params_shard_26.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.31.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "8079166d7d7e7e2e93848235373bd1a2" }, { "dataPath": "params_shard_27.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.31.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.31.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.31.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.31.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.31.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.31.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "59639205b28fc6405fd0fb95d1e14867" }, { "dataPath": "params_shard_28.bin", "format": "raw-shard", "nbytes": 49250304, "records": [ { "name": "transformer.embd.q_weight", "shape": [ 32064, 384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 49250304, "byteOffset": 0 } ], "md5sum": "e61f1b06e77de108c4a56240a97fbb3b" }, { "dataPath": "params_shard_29.bin", "format": "raw-shard", "nbytes": 22093824, "records": [ { "name": "transformer.h.31.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.31.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.norm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.embd.q_scale", "shape": [ 32064, 96 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6156288, "byteOffset": 15931392 }, { "name": "transformer.h.0.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 22087680 } ], "md5sum": "cb39bbac8ff3fbd25f095b151fc7d353" }, { "dataPath": "params_shard_30.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.0.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "8053ff98adff83cdb53acbbd266fb222" }, { "dataPath": "params_shard_31.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.0.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.0.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.0.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.0.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.0.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.0.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "e1c292712e5883d409ebf8d8ccfafc64" }, { "dataPath": "params_shard_32.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.1.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "b8c128ed7c6b1b42cc87788bc3fc7f6c" }, { "dataPath": "params_shard_33.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.0.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.0.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.1.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.1.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.1.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.1.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.1.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "20140e9d4ffa8665cecbf0ed1ed89171" }, { "dataPath": "params_shard_34.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.1.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.1.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.1.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.1.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.10.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "c5778b22d0c307954d92db51a781348c" }, { "dataPath": "params_shard_35.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.10.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "9ce87cf585d4c6fb8e2a7e3b49289714" }, { "dataPath": "params_shard_36.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.10.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.10.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.10.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.10.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.10.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.10.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "7ea31f0f3d66bca695addd7789e60c1b" }, { "dataPath": "params_shard_37.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.11.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "555092402acc1e13d8f45466375f0747" }, { "dataPath": "params_shard_38.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.10.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.10.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.11.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.11.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.11.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.11.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.11.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "c1215c69edf03326f3dea4452f7fa761" }, { "dataPath": "params_shard_39.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.11.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.11.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.11.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.11.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.12.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "9bc10add79a9f2a324b8e18671cc09bd" }, { "dataPath": "params_shard_40.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.12.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "0a897ff9bd8bbe24b8c8a6c0a3ebd8b7" }, { "dataPath": "params_shard_41.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.12.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.12.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.12.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.12.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.12.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.12.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "7aaab1972cc0f3d393e96aaab12edd65" }, { "dataPath": "params_shard_42.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.13.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "42604cc962f6377027b06eb726972840" }, { "dataPath": "params_shard_43.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.12.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.12.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.13.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.13.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.13.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.13.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.13.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "183e00f45855a504b51ab3611582ebe5" }, { "dataPath": "params_shard_44.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.13.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.13.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.13.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.13.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.14.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "5393801fd565cd3f118c0a74b162b5d6" }, { "dataPath": "params_shard_45.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.14.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "0888063bb3e5ed35275876502d983288" }, { "dataPath": "params_shard_46.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.14.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.14.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.14.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.14.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.14.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.14.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "f48e479383c2f129d4d831532828ec90" }, { "dataPath": "params_shard_47.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.15.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "c7751ffa8b059fe61f9ad0bc45177302" }, { "dataPath": "params_shard_48.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.14.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.14.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.15.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.15.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.15.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.15.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.15.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "a609dc5b6dbf8487bb3bee5b7731e7c6" }, { "dataPath": "params_shard_49.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.15.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.15.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.15.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.15.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.16.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "deb03e306c04aec3d568a5eeaa5546ff" }, { "dataPath": "params_shard_50.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.16.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "e88c4e1771cfe22c6cc0711147bad793" }, { "dataPath": "params_shard_51.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.16.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.16.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.16.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.16.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.16.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.16.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "11f09ca4dbaaaade468299c0f840bcd3" }, { "dataPath": "params_shard_52.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.17.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "9e64bf8d0b4bff6f81865a6cfefceb9e" }, { "dataPath": "params_shard_53.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.16.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.16.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.17.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.17.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.17.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.17.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.17.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "a1cff643fb5621e707b8ffbe44bd0bf3" }, { "dataPath": "params_shard_54.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.17.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.17.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.17.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.17.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.18.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "09a90e1f53fcae525648794510cf1fbc" }, { "dataPath": "params_shard_55.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.18.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "922b073785a159480d5fc79f94dc50c9" }, { "dataPath": "params_shard_56.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.18.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.18.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.18.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.18.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.18.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.18.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "7b0496a486806f50a0e59c3a1530f941" }, { "dataPath": "params_shard_57.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.19.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "3217ce87180b1316ade0c0b0504a7526" }, { "dataPath": "params_shard_58.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.18.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.18.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.19.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.19.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.19.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.19.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.19.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "6cdcce24eb1c8f4f4acfef2bf9df4119" }, { "dataPath": "params_shard_59.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.19.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.19.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.19.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.19.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.2.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "b5bdccd591c2210ea01f1a55a49fdd6a" }, { "dataPath": "params_shard_60.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.2.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "4591d8d6b372ca9210cd7078cbd12d03" }, { "dataPath": "params_shard_61.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.2.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.2.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.2.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.2.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.2.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.2.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "744ed8be520511f8d118f51ce553a759" }, { "dataPath": "params_shard_62.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.20.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "5e5990c1bb2ac54f8c36061fe3e88a1b" }, { "dataPath": "params_shard_63.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.2.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.2.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.20.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.20.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.20.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.20.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.20.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "372a721c75d4bc4270b132ca62a4f668" }, { "dataPath": "params_shard_64.bin", "format": "raw-shard", "nbytes": 26548224, "records": [ { "name": "transformer.h.20.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.20.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.20.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.20.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.21.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 21233664 }, { "name": "transformer.h.21.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 25952256 }, { "name": "transformer.h.3.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 26542080 } ], "md5sum": "8817389194f45a402af91bdbcb4ff75d" }, { "dataPath": "params_shard_65.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.3.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "81f5e9b8a5414eca83aa828d9df8b77f" }, { "dataPath": "params_shard_66.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.3.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.3.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.3.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.3.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.3.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.3.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "779d1a6fae10f8f87a456e09345d85cb" }, { "dataPath": "params_shard_67.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.4.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "f665b53505572c15e4e6fc0fac7360a9" }, { "dataPath": "params_shard_68.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.3.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.3.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.4.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.4.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.4.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.4.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.4.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "6108f7953bc61acc85abad6628e2fb42" }, { "dataPath": "params_shard_69.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.4.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.4.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.4.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.4.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.5.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "cb087ffffb00b1755b205047dd403838" }, { "dataPath": "params_shard_70.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.5.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "c406e3d3d04e6d0c3a6a2ebd4fe474fe" }, { "dataPath": "params_shard_71.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.5.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.5.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.5.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.5.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.5.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.5.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "67418d6b1fbf27b1befb89a688a2953b" }, { "dataPath": "params_shard_72.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.6.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "c8082b14303b93a635fbfe1641bcfe19" }, { "dataPath": "params_shard_73.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.5.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.5.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.6.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.6.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.6.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.6.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.6.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "44781bdf39e208a53b257ce73725da3a" }, { "dataPath": "params_shard_74.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.6.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.6.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.6.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.6.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.7.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "886645852ef2590ae080d36982d8d8d7" }, { "dataPath": "params_shard_75.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.7.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "e00ed4d771be5bb0f790c08cf2a92e47" }, { "dataPath": "params_shard_76.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.7.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.7.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.7.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.7.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.7.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.7.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "3b7d62d668c9e2db47d49e36cfd97d5c" }, { "dataPath": "params_shard_77.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.8.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "8270eebb6525355e6b89451d075097b1" }, { "dataPath": "params_shard_78.bin", "format": "raw-shard", "nbytes": 33239040, "records": [ { "name": "transformer.h.7.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.7.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 }, { "name": "transformer.h.8.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 15925248 }, { "name": "transformer.h.8.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 15931392 }, { "name": "transformer.h.8.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 28514304 }, { "name": "transformer.h.8.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 30087168 }, { "name": "transformer.h.8.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 33232896 } ], "md5sum": "3d02821b738f5ecc72223d132ef37fc9" }, { "dataPath": "params_shard_79.bin", "format": "raw-shard", "nbytes": 21239808, "records": [ { "name": "transformer.h.8.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 0 }, { "name": "transformer.h.8.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 4718592 }, { "name": "transformer.h.8.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 5308416 }, { "name": "transformer.h.8.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 19464192 }, { "name": "transformer.h.9.ln.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 21233664 } ], "md5sum": "691e55b2c31bbb89efc566e608e11a51" }, { "dataPath": "params_shard_80.bin", "format": "raw-shard", "nbytes": 25165824, "records": [ { "name": "transformer.h.9.mlp.gate_up_proj.q_weight", "shape": [ 384, 16384 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 25165824, "byteOffset": 0 } ], "md5sum": "e9d16ab85829a080c4dbc99c69bd4505" }, { "dataPath": "params_shard_81.bin", "format": "raw-shard", "nbytes": 22616064, "records": [ { "name": "transformer.h.9.mlp.down_proj.q_weight", "shape": [ 1024, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 12582912, "byteOffset": 0 }, { "name": "transformer.h.9.mlp.down_proj.q_scale", "shape": [ 256, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1572864, "byteOffset": 12582912 }, { "name": "transformer.h.9.mlp.gate_up_proj.q_scale", "shape": [ 96, 16384 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 3145728, "byteOffset": 14155776 }, { "name": "transformer.h.9.post_attention_layernorm.weight", "shape": [ 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 6144, "byteOffset": 17301504 }, { "name": "transformer.h.9.mixer.out_proj.q_weight", "shape": [ 384, 3072 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 4718592, "byteOffset": 17307648 }, { "name": "transformer.h.9.mixer.out_proj.q_scale", "shape": [ 96, 3072 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 589824, "byteOffset": 22026240 } ], "md5sum": "9dc47daa5be5c62b51b14abffefb9f2a" }, { "dataPath": "params_shard_82.bin", "format": "raw-shard", "nbytes": 15925248, "records": [ { "name": "transformer.h.9.mixer.qkv_proj.q_weight", "shape": [ 384, 9216 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 14155776, "byteOffset": 0 }, { "name": "transformer.h.9.mixer.qkv_proj.q_scale", "shape": [ 96, 9216 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1769472, "byteOffset": 14155776 } ], "md5sum": "9ed7291a49ebc6209ab1e5ddd7b35ac0" } ] }