{
"metadata": {
"ParamSize": 99,
"ParamBytes": 9080938496.0,
"BitsPerParam": 16.0
},
"records": [
{
"dataPath": "params_shard_0.bin",
"format": "raw-shard",
"nbytes": 1050673152,
"records": [
{
"name": "lm_head.weight",
"shape": [
128256,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 1050673152,
"byteOffset": 0
}
],
"md5sum": "75edd591cd36aeac9b51ef921ce49e6d"
},
{
"dataPath": "params_shard_1.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.10.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "882adac81ef5715c534ae6b6dfca6a8a"
},
{
"dataPath": "params_shard_2.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.10.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "6685f6e417a8f6032e1e648685682421"
},
{
"dataPath": "params_shard_3.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.10.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "bb0d281ced9faf26f31bef662e487446"
},
{
"dataPath": "params_shard_4.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.10.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "c64e84066f502e276d6e632d89deb55e"
},
{
"dataPath": "params_shard_5.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.11.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "53e53d7d42ea7237d58db0c90207f321"
},
{
"dataPath": "params_shard_6.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.11.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "b4cafce0dd2ad86578f5ef0c598d767d"
},
{
"dataPath": "params_shard_7.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.11.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "0ab8a1a37b3edf246e610db0b820407e"
},
{
"dataPath": "params_shard_8.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.11.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "559c69b36e2dcb7288641032f85176a8"
},
{
"dataPath": "params_shard_9.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.12.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "74ace719a6dbb05c2eba2beaf4066d72"
},
{
"dataPath": "params_shard_10.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.12.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "aeef8a609db53e108557bad1f10a3be7"
},
{
"dataPath": "params_shard_11.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.12.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "9ff3eca975f628c98f48541c57fddf3b"
},
{
"dataPath": "params_shard_12.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.12.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "41e28bb232e53fa15619a7103ef5e0d6"
},
{
"dataPath": "params_shard_13.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.13.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "6e75989b9677586f21f73591ae348135"
},
{
"dataPath": "params_shard_14.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.13.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "1ca35e693290541b8b9ba1d149e7e429"
},
{
"dataPath": "params_shard_15.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.13.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "fbbd78f04c1a0ac7dc7801fd432f0cfc"
},
{
"dataPath": "params_shard_16.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.13.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "069e926e8a015214466100ff3e6ff924"
},
{
"dataPath": "params_shard_17.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.14.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "36913890da219b82c0535faa63141777"
},
{
"dataPath": "params_shard_18.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.14.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "83fec9c67c69936a8eb7efab2edd755d"
},
{
"dataPath": "params_shard_19.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.14.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "8095fd912c1622273c11fd2a971d244c"
},
{
"dataPath": "params_shard_20.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.14.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "af1902028e36c5819e52592072ce67a5"
},
{
"dataPath": "params_shard_21.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.15.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "8189251b9267ac10b1df23d471730c45"
},
{
"dataPath": "params_shard_22.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.15.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "7f1509e4cdc899ef0472fbeac943906a"
},
{
"dataPath": "params_shard_23.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.15.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "43518c0a6fff6447442a6d89efa1ff50"
},
{
"dataPath": "params_shard_24.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.15.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "ea9cf82232756ec2834bbdaa38b243c2"
},
{
"dataPath": "params_shard_25.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.9.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "55b04d0c593293dadb6f08db1bed6755"
},
{
"dataPath": "params_shard_26.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.9.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "242c8c0fdbee5a54f4d1453f66a2b9cf"
},
{
"dataPath": "params_shard_27.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.9.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "169e455da07f15b7bfd0f86b9e13920d"
},
{
"dataPath": "params_shard_28.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.9.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "9dbb023c23e0b6b9b095eca81e6e6ffc"
},
{
"dataPath": "params_shard_29.bin",
"format": "raw-shard",
"nbytes": 1050673152,
"records": [
{
"name": "model.embed_tokens.weight",
"shape": [
128256,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 1050673152,
"byteOffset": 0
}
],
"md5sum": "61a3075c1d56322e3b21589b5a40077a"
},
{
"dataPath": "params_shard_30.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.0.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "e38adc5f6cd900d6a4ddce51c418b621"
},
{
"dataPath": "params_shard_31.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.0.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "2688b8c753a249db64bb8096cbb4ee46"
},
{
"dataPath": "params_shard_32.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.0.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "bde37a0116316bc7a2edb954db041f9a"
},
{
"dataPath": "params_shard_33.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.0.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "0751d8c096fb83e2449ad595823c329d"
},
{
"dataPath": "params_shard_34.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.1.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "01b027498d27b92bfcc84183e5d3d34c"
},
{
"dataPath": "params_shard_35.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.1.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "47d49cf7c227f7486cad873bedebcc35"
},
{
"dataPath": "params_shard_36.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.1.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "6c6ce14ce180475895eec3d83946c1cb"
},
{
"dataPath": "params_shard_37.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.1.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "89605654f707dbe9ad46f348065b2a30"
},
{
"dataPath": "params_shard_38.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.2.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "36be3875e892660f5ce0130d3cf0e123"
},
{
"dataPath": "params_shard_39.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.2.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "cbaebeea0c7212f5c3d155d623c373a4"
},
{
"dataPath": "params_shard_40.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.2.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "a1d8e1d654e66a793c9570296a38120d"
},
{
"dataPath": "params_shard_41.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.2.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "bfb9e277e026e401401d1f2ff74b0bb4"
},
{
"dataPath": "params_shard_42.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.3.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "8ba44f0035b36f8400d8d58e2b1048cf"
},
{
"dataPath": "params_shard_43.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.3.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "59d3f67b229b77d3c7d636fa81956d13"
},
{
"dataPath": "params_shard_44.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.3.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "2b60c35074e68fe4c2c9d2fb39411bcf"
},
{
"dataPath": "params_shard_45.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.3.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "fd0947763bd4a6c920df8f115d281b59"
},
{
"dataPath": "params_shard_46.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.4.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "c0515ff9e26c249190d472180cfd75fa"
},
{
"dataPath": "params_shard_47.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.4.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "74951cccccb4365035c8e64db27a0153"
},
{
"dataPath": "params_shard_48.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.4.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "f5ac4ef990bdadbaa941f1d3fa38a0b9"
},
{
"dataPath": "params_shard_49.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.4.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "9036a0a1d92841dd1720590232ff092b"
},
{
"dataPath": "params_shard_50.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.5.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "bbab290a1a40b518721a5c287f313e22"
},
{
"dataPath": "params_shard_51.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.5.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "fdb9cf70c17fec072633c03e4cf0b63e"
},
{
"dataPath": "params_shard_52.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.5.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "9693c2aa897188396cd3eab7040d51ce"
},
{
"dataPath": "params_shard_53.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.5.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "8334828cca96b690fbc2c612ec5d06ea"
},
{
"dataPath": "params_shard_54.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.6.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "783b2d84e42a6402397001749439c350"
},
{
"dataPath": "params_shard_55.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.6.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "68c0820b730f9b61fe0390ec719c4e45"
},
{
"dataPath": "params_shard_56.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.6.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "4a39640abc08329a9b19c2a847bc461a"
},
{
"dataPath": "params_shard_57.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.6.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "2636ab38f2bdf0693f8672e9b23f5927"
},
{
"dataPath": "params_shard_58.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.7.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "31d16ba9a1b8044311b6da2fb1f802e3"
},
{
"dataPath": "params_shard_59.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.7.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "04a2c934711cd4ca9c06256add28c53d"
},
{
"dataPath": "params_shard_60.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.7.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "29c78d4b12bf9dd955a8e474b0a95560"
},
{
"dataPath": "params_shard_61.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.7.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "719fb7aae11d67061f7e658fc85d588f"
},
{
"dataPath": "params_shard_62.bin",
"format": "raw-shard",
"nbytes": 117440512,
"records": [
{
"name": "model.layers.8.mlp.down_proj.weight",
"shape": [
4096,
14336
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 117440512,
"byteOffset": 0
}
],
"md5sum": "741a1aaadd2800d2d7c8d53dbc419704"
},
{
"dataPath": "params_shard_63.bin",
"format": "raw-shard",
"nbytes": 234881024,
"records": [
{
"name": "model.layers.8.mlp.gate_up_proj.weight",
"shape": [
28672,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 234881024,
"byteOffset": 0
}
],
"md5sum": "6a2bb53b5592bc8a752e2a53d2ebb556"
},
{
"dataPath": "params_shard_64.bin",
"format": "raw-shard",
"nbytes": 50331648,
"records": [
{
"name": "model.layers.8.self_attn.qkv_proj.weight",
"shape": [
6144,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 50331648,
"byteOffset": 0
}
],
"md5sum": "69d787685687a5273cb624f8062a1607"
},
{
"dataPath": "params_shard_65.bin",
"format": "raw-shard",
"nbytes": 33554432,
"records": [
{
"name": "model.layers.8.self_attn.o_proj.weight",
"shape": [
4096,
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 33554432,
"byteOffset": 0
}
],
"md5sum": "63e433d6f4cec793a14223c706b9d3d9"
},
{
"dataPath": "params_shard_66.bin",
"format": "raw-shard",
"nbytes": 270336,
"records": [
{
"name": "model.layers.10.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 0
},
{
"name": "model.layers.10.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 8192
},
{
"name": "model.layers.11.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 16384
},
{
"name": "model.layers.11.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 24576
},
{
"name": "model.layers.12.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 32768
},
{
"name": "model.layers.12.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 40960
},
{
"name": "model.layers.13.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 49152
},
{
"name": "model.layers.13.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 57344
},
{
"name": "model.layers.14.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 65536
},
{
"name": "model.layers.14.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 73728
},
{
"name": "model.layers.15.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 81920
},
{
"name": "model.layers.15.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 90112
},
{
"name": "model.layers.9.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 98304
},
{
"name": "model.layers.9.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 106496
},
{
"name": "model.norm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 114688
},
{
"name": "model.layers.0.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 122880
},
{
"name": "model.layers.0.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 131072
},
{
"name": "model.layers.1.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 139264
},
{
"name": "model.layers.1.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 147456
},
{
"name": "model.layers.2.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 155648
},
{
"name": "model.layers.2.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 163840
},
{
"name": "model.layers.3.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 172032
},
{
"name": "model.layers.3.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 180224
},
{
"name": "model.layers.4.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 188416
},
{
"name": "model.layers.4.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 196608
},
{
"name": "model.layers.5.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 204800
},
{
"name": "model.layers.5.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 212992
},
{
"name": "model.layers.6.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 221184
},
{
"name": "model.layers.6.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 229376
},
{
"name": "model.layers.7.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 237568
},
{
"name": "model.layers.7.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 245760
},
{
"name": "model.layers.8.input_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 253952
},
{
"name": "model.layers.8.post_attention_layernorm.weight",
"shape": [
4096
],
"dtype": "float16",
"format": "f32-to-bf16",
"nbytes": 8192,
"byteOffset": 262144
}
],
"md5sum": "02d3d0ecea966763230cf838b3702024"
}
]
}