{ "metadata": { "ParamSize": 455, "ParamBytes": 1740533760.0, "BitsPerParam": 5.003107184052364 }, "records": [ { "dataPath": "params_shard_0.bin", "format": "raw-shard", "nbytes": 65536000, "records": [ { "name": "lm_head.linear.q_weight", "shape": [ 51200, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 65536000, "byteOffset": 0 } ], "md5sum": "9c19001bd196fd811aaab4570dab83c7" }, { "dataPath": "params_shard_1.bin", "format": "raw-shard", "nbytes": 23091200, "records": [ { "name": "lm_head.linear.bias", "shape": [ 51200 ], "dtype": "bfloat16", "format": "raw", "nbytes": 102400, "byteOffset": 0 }, { "name": "lm_head.linear.q_scale", "shape": [ 51200, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 8192000, "byteOffset": 102400 }, { "name": "lm_head.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 8294400 }, { "name": "lm_head.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 8299520 }, { "name": "transformer.h.30.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 8304640 }, { "name": "transformer.h.30.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 8320000 }, { "name": "transformer.h.30.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 18150400 }, { "name": "transformer.h.30.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 19379200 }, { "name": "transformer.h.30.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 19384320 }, { "name": "transformer.h.30.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 22661120 }, { "name": "transformer.h.30.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 23070720 } ], "md5sum": "70f879f4f894df6c33804481f8c9d18c" }, { "dataPath": "params_shard_2.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.30.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.30.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.30.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.30.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.30.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.31.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.31.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.31.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "96b429155ddeff3bd12e4224035110ce" }, { "dataPath": "params_shard_3.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.31.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.31.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.31.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.31.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.31.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.31.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.31.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.31.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.31.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "0da0fbbf609ba06f26fe27df8c9f68f5" }, { "dataPath": "params_shard_4.bin", "format": "raw-shard", "nbytes": 65536000, "records": [ { "name": "transformer.embd.q_weight", "shape": [ 51200, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 65536000, "byteOffset": 0 } ], "md5sum": "cd2ab4b418c2cd5fbd240da8d981f1d8" }, { "dataPath": "params_shard_5.bin", "format": "raw-shard", "nbytes": 32793600, "records": [ { "name": "transformer.h.31.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.31.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.embd.q_scale", "shape": [ 51200, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 8192000, "byteOffset": 14745600 }, { "name": "transformer.h.0.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 22937600 }, { "name": "transformer.h.0.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 22942720 }, { "name": "transformer.h.0.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 22947840 }, { "name": "transformer.h.0.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 22963200 } ], "md5sum": "cd751b1232b8c4787461a0ee2e2491f1" }, { "dataPath": "params_shard_6.bin", "format": "raw-shard", "nbytes": 32798720, "records": [ { "name": "transformer.h.0.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 0 }, { "name": "transformer.h.0.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 1228800 }, { "name": "transformer.h.0.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 1233920 }, { "name": "transformer.h.0.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 4510720 }, { "name": "transformer.h.0.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 4920320 }, { "name": "transformer.h.0.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 4940800 }, { "name": "transformer.h.0.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 18048000 }, { "name": "transformer.h.0.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 19686400 }, { "name": "transformer.h.0.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19691520 } ], "md5sum": "d4a305ba2fb8c512bef3ccec7cb44f73" }, { "dataPath": "params_shard_7.bin", "format": "raw-shard", "nbytes": 31185920, "records": [ { "name": "transformer.h.0.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.1.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 1638400 }, { "name": "transformer.h.1.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 1643520 }, { "name": "transformer.h.1.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 1648640 }, { "name": "transformer.h.1.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 1664000 }, { "name": "transformer.h.1.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 11494400 }, { "name": "transformer.h.1.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 12723200 }, { "name": "transformer.h.1.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 12728320 }, { "name": "transformer.h.1.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 16005120 }, { "name": "transformer.h.1.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 16414720 }, { "name": "transformer.h.1.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 16435200 }, { "name": "transformer.h.1.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 29542400 }, { "name": "transformer.h.1.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 31180800 } ], "md5sum": "379ba25c7f60f7033a67ab7be685f2c3" }, { "dataPath": "params_shard_8.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.1.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.1.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.10.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.10.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.10.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.10.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.10.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.10.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.10.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.10.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.10.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "0ec2f1c958dbbbc3d71ecc28a51a16ca" }, { "dataPath": "params_shard_9.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.10.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.10.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.10.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.10.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.10.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.11.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.11.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.11.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "3d450ce64db570ae1c449cf544acb16f" }, { "dataPath": "params_shard_10.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.11.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.11.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.11.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.11.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.11.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.11.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.11.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.11.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.11.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "008d3e1e51de2db3d12cac57f19a1473" }, { "dataPath": "params_shard_11.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.11.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.11.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.12.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.12.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.12.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.12.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.12.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.12.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.12.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.12.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.12.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "5fd181b12bd4fef6a14210104005985f" }, { "dataPath": "params_shard_12.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.12.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.12.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.12.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.12.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.12.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.13.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.13.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.13.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "70f8fa230cbf9fa0605ad2bfe50eda08" }, { "dataPath": "params_shard_13.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.13.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.13.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.13.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.13.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.13.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.13.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.13.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.13.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.13.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "6d678ae5acdae8c517b234604934f4fa" }, { "dataPath": "params_shard_14.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.13.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.13.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.14.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.14.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.14.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.14.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.14.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.14.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.14.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.14.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.14.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "9b6800be2a6b408875831e10146759dd" }, { "dataPath": "params_shard_15.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.14.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.14.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.14.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.14.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.14.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.15.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.15.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.15.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "cbdb75a8458c462beb74abf63b002994" }, { "dataPath": "params_shard_16.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.15.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.15.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.15.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.15.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.15.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.15.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.15.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.15.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.15.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "53d7d6a7303145851cfead9e3e5d2cd9" }, { "dataPath": "params_shard_17.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.15.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.15.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.16.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.16.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.16.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.16.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.16.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.16.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.16.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.16.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.16.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "a0497b03102aea0f572520068d63a0a4" }, { "dataPath": "params_shard_18.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.16.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.16.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.16.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.16.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.16.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.17.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.17.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.17.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "7fadb86755d756942a44c0c9907e5291" }, { "dataPath": "params_shard_19.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.17.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.17.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.17.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.17.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.17.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.17.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.17.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.17.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.17.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "26c11e60dc27a5a8dffa5856fd90cfb7" }, { "dataPath": "params_shard_20.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.17.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.17.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.18.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.18.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.18.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.18.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.18.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.18.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.18.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.18.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.18.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "e31917c5e35e2372f4daa31531c2eb13" }, { "dataPath": "params_shard_21.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.18.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.18.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.18.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.18.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.18.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.19.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.19.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.19.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "e03087ffa3fcc27b5161fbf6b9578517" }, { "dataPath": "params_shard_22.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.19.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.19.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.19.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.19.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.19.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.19.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.19.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.19.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.19.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "3acc8adbd6e813d9d1f60439be3b0885" }, { "dataPath": "params_shard_23.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.19.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.19.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.2.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.2.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.2.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.2.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.2.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.2.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.2.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.2.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.2.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "ccf7506742d97cbd532ecb0617dcec36" }, { "dataPath": "params_shard_24.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.2.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.2.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.2.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.2.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.2.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.20.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.20.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.20.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "7beda28ef0992ad643e5d92efdb23634" }, { "dataPath": "params_shard_25.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.20.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.20.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.20.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.20.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.20.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.20.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.20.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.20.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.20.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "1590cd7f8fca3b6b4bbc2a304af04494" }, { "dataPath": "params_shard_26.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.20.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.20.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.21.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.21.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.21.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.21.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.21.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.21.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.21.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.21.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.21.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "482ab177d9fb9c3f4ab1e1cf05ae175e" }, { "dataPath": "params_shard_27.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.21.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.21.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.21.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.21.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.21.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.22.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.22.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.22.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "262f8afa074bb21c3ea4c49344036a0a" }, { "dataPath": "params_shard_28.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.22.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.22.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.22.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.22.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.22.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.22.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.22.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.22.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.22.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "3afd49706c6bf92118dabaa1796ea52e" }, { "dataPath": "params_shard_29.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.22.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.22.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.23.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.23.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.23.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.23.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.23.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.23.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.23.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.23.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.23.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "b4419b49a7cc7c42bb62fd8b56f3327a" }, { "dataPath": "params_shard_30.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.23.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.23.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.23.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.23.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.23.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.24.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.24.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.24.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "480ea4aeb812c503bacebee6c3ff7a7d" }, { "dataPath": "params_shard_31.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.24.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.24.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.24.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.24.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.24.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.24.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.24.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.24.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.24.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "078411d50dd35462c256b8de369ca070" }, { "dataPath": "params_shard_32.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.24.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.24.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.25.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.25.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.25.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.25.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.25.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.25.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.25.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.25.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.25.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "61516932f0a10f1b60935225aba2f267" }, { "dataPath": "params_shard_33.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.25.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.25.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.25.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.25.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.25.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.26.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.26.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.26.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "dc2099f4fd4ba1e7b08e7ce045987fed" }, { "dataPath": "params_shard_34.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.26.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.26.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.26.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.26.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.26.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.26.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.26.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.26.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.26.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "c66549e2e4f5075435eb73f55d60c6ee" }, { "dataPath": "params_shard_35.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.26.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.26.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.27.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.27.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.27.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.27.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.27.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.27.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.27.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.27.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.27.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "1074749621c045ebbe59d1c39ddcb738" }, { "dataPath": "params_shard_36.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.27.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.27.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.27.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.27.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.27.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.28.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.28.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.28.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "8157b613091de9dfccde627b2a844ce7" }, { "dataPath": "params_shard_37.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.28.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.28.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.28.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.28.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.28.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.28.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.28.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.28.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.28.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "9f17118ce1175b9a606142009144b9c6" }, { "dataPath": "params_shard_38.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.28.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.28.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.29.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.29.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.29.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.29.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.29.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.29.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.29.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.29.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.29.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "c7ec0b9ca6a3f1231930f561aa1e6398" }, { "dataPath": "params_shard_39.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.29.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.29.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.29.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.29.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.29.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.3.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.3.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.3.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "adf21882043b41edd079b3f6b0682d3c" }, { "dataPath": "params_shard_40.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.3.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.3.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.3.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.3.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.3.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.3.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.3.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.3.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.3.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "711bcfc16fc5b1df4290821d4b97298a" }, { "dataPath": "params_shard_41.bin", "format": "raw-shard", "nbytes": 29552640, "records": [ { "name": "transformer.h.3.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.3.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.30.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.30.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.4.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14755840 }, { "name": "transformer.h.4.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14760960 }, { "name": "transformer.h.4.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14766080 }, { "name": "transformer.h.4.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14781440 }, { "name": "transformer.h.4.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24611840 }, { "name": "transformer.h.4.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25840640 }, { "name": "transformer.h.4.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25845760 }, { "name": "transformer.h.4.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29122560 }, { "name": "transformer.h.4.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29532160 } ], "md5sum": "548842d1dfb8df8c4e7e1766729a665b" }, { "dataPath": "params_shard_42.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.4.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.4.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.4.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.4.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.4.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.5.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.5.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.5.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "ead9fc15f9023d950a5ee955fe9e9902" }, { "dataPath": "params_shard_43.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.5.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.5.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.5.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.5.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.5.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.5.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.5.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.5.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.5.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "8aea60bc23a32d99efe6cd1c01a7ba8d" }, { "dataPath": "params_shard_44.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.5.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.5.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.6.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.6.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.6.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.6.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.6.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.6.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.6.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.6.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.6.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "6200d2d13f9c0e9231439d9ed58e1c6b" }, { "dataPath": "params_shard_45.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.6.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.6.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.6.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.6.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.6.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.7.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.7.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.7.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "19c90209665e06b9b9b4bb0748f7e8f8" }, { "dataPath": "params_shard_46.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.7.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.7.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.7.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.7.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.7.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.7.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.7.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.7.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.7.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "1146ad1b0594857dbe1426077997cdfc" }, { "dataPath": "params_shard_47.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.7.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.7.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.8.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.8.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.8.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.8.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.8.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.8.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.8.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.8.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.8.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "7e27407c720579f19b80db965f08144d" }, { "dataPath": "params_shard_48.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.8.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.8.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.8.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.8.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.8.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.9.ln.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.9.ln.weight", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.9.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "bfloat16", "format": "raw", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "d0e1ba04100f45e25fa640fb5230794a" }, { "dataPath": "params_shard_49.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.9.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.9.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.9.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.9.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.9.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.9.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "bfloat16", "format": "raw", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.9.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.9.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.9.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "bfloat16", "format": "raw", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "680cdef2e057f4941af38e1b066e9fc2" }, { "dataPath": "params_shard_50.bin", "format": "raw-shard", "nbytes": 14745600, "records": [ { "name": "transformer.h.9.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.9.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "bfloat16", "format": "raw", "nbytes": 1638400, "byteOffset": 13107200 } ], "md5sum": "6400f7aefd9a12dc774df07ab0937b4c" } ] }