{ "metadata": { "ParamSize": 405, "ParamBytes": 8246568960.0, "BitsPerParam": 4.500325336993593 }, "records": [ { "dataPath": "params_shard_0.bin", "format": "raw-shard", "nbytes": 256901120, "records": [ { "name": "lm_head.q_weight", "shape": [ 100352, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 256901120, "byteOffset": 0 } ], "md5sum": "77f2d0330d90f43eba8f26efea0421aa" }, { "dataPath": "params_shard_1.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.34.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "cba03535921c817e48c87561cb1552ca" }, { "dataPath": "params_shard_2.bin", "format": "raw-shard", "nbytes": 32122880, "records": [ { "name": "lm_head.q_scale", "shape": [ 100352, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 32112640, "byteOffset": 0 }, { "name": "transformer.h.34.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 32112640 } ], "md5sum": "381742e9c421cc7f97992d29c741d6ac" }, { "dataPath": "params_shard_3.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.34.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "d50a0da9e92c1ec8be71c091781f5ee1" }, { "dataPath": "params_shard_4.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.35.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "47751759edabeb1b6c3ff2490cfa4c14" }, { "dataPath": "params_shard_5.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.35.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "80017fedf3d07b8f507736270bf7115f" }, { "dataPath": "params_shard_6.bin", "format": "raw-shard", "nbytes": 22958080, "records": [ { "name": "transformer.h.34.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.34.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.34.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.35.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17213440 }, { "name": "transformer.h.35.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 17223680 } ], "md5sum": "2b0f44a59f8e75509a9cfb787cd11042" }, { "dataPath": "params_shard_7.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.35.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "e21a91b5b675e393d8d119e2d73fcbe4" }, { "dataPath": "params_shard_8.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.36.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "a339b875f7549c7cad3828aef87566bb" }, { "dataPath": "params_shard_9.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.35.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.35.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.35.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.35.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.35.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.36.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "4eadc0f8465b0a0a021380829cd9feaf" }, { "dataPath": "params_shard_10.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.36.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "77379218f3817c3863304e3538d851ff" }, { "dataPath": "params_shard_11.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.36.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "1fe22d5ee37de780b3c0dfb0a7c334c5" }, { "dataPath": "params_shard_12.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.36.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.36.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.36.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.36.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.36.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "b831eb02e7636ccfdd2e7e309bd9abfd" }, { "dataPath": "params_shard_13.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.37.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "d3a81755352ded55633baa549c28bd86" }, { "dataPath": "params_shard_14.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.37.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "17dd3e58d82b38bd91d2ccf1bfa95261" }, { "dataPath": "params_shard_15.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.36.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.37.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.37.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.37.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.37.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.37.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "2b4c8ce2d86610e4866a8354b6b9afb7" }, { "dataPath": "params_shard_16.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.38.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "be0bcbb026c4b13b21823394e7465781" }, { "dataPath": "params_shard_17.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.38.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "ee65e2059282b4275af35f73266f2173" }, { "dataPath": "params_shard_18.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.37.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.37.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.37.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.38.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.38.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "4d596fc62234668ecfc8a8894c44577a" }, { "dataPath": "params_shard_19.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.38.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "5cfb730205a3a5a04c594050957d3c3d" }, { "dataPath": "params_shard_20.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.39.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "f1cb1e1126c68c3c3baa187738e2ee81" }, { "dataPath": "params_shard_21.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.38.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.38.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.38.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.38.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.38.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.39.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "a6282c5e80536e2a9eb3869d814fe8f2" }, { "dataPath": "params_shard_22.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.39.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "11ec517dc5e8b7206aaa2f1b7aea763e" }, { "dataPath": "params_shard_23.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.39.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "d931d979c902a24d5aebf617b8764a45" }, { "dataPath": "params_shard_24.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.39.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.39.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.39.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.39.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.39.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "eb3e246736e2e22405fb90529e776c7a" }, { "dataPath": "params_shard_25.bin", "format": "raw-shard", "nbytes": 256901120, "records": [ { "name": "transformer.embd.q_weight", "shape": [ 100352, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 256901120, "byteOffset": 0 } ], "md5sum": "e86c750e94ff4ffc984137ddfebe77eb" }, { "dataPath": "params_shard_26.bin", "format": "raw-shard", "nbytes": 32112640, "records": [ { "name": "transformer.embd.q_scale", "shape": [ 100352, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 32112640, "byteOffset": 0 } ], "md5sum": "cf0eed4010bda4aa4cf8fe6e19288d36" }, { "dataPath": "params_shard_27.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.0.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "9c1c4dae8d0241e26c9b94fe46660894" }, { "dataPath": "params_shard_28.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.0.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "f6329b773a14db054b800a6aa6985afc" }, { "dataPath": "params_shard_29.bin", "format": "raw-shard", "nbytes": 32798720, "records": [ { "name": "transformer.h.39.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.norm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.0.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2467840 }, { "name": "transformer.h.0.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2478080 }, { "name": "transformer.h.0.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8212480 }, { "name": "transformer.h.0.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19681280 }, { "name": "transformer.h.0.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19691520 } ], "md5sum": "91cba26e80bffa55ebf0348d5aeb9b57" }, { "dataPath": "params_shard_30.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.1.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "7f1a3177b998db14a386508a3555f420" }, { "dataPath": "params_shard_31.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.1.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "2bfffc8a1b1d87ccec33b014da4f0a56" }, { "dataPath": "params_shard_32.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.0.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.0.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.0.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.1.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.1.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "f069d71741b10abe70feb4b841a0015e" }, { "dataPath": "params_shard_33.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.1.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "69e887495da60c9a88d6e51d3f2911a3" }, { "dataPath": "params_shard_34.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.2.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "63bab60ce07a70a94d03867f7768e123" }, { "dataPath": "params_shard_35.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.1.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.1.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.1.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.1.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.1.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.2.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "b37730f25b880aa09e8136fbb4ccc93c" }, { "dataPath": "params_shard_36.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.2.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "75e610582e0d23c4e7b17a9b8cb38fb2" }, { "dataPath": "params_shard_37.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.2.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "0d1b0aa52c2a38a492cb7e9795c7ca51" }, { "dataPath": "params_shard_38.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.2.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.2.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.2.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.2.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.2.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "803ee2d942b55bf426e535cb565536d9" }, { "dataPath": "params_shard_39.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.3.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "6a1120fc4084fe08b64a561c1a5da733" }, { "dataPath": "params_shard_40.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.3.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "b6730295f6fe17e37f2a211639f8ade8" }, { "dataPath": "params_shard_41.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.2.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.3.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.3.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.3.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.3.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.3.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "d2a1535a1eb8b8e97ccea7a7b5da8a5e" }, { "dataPath": "params_shard_42.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.4.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "53de62cc0107a4ed6f4a9ab5e9416e4a" }, { "dataPath": "params_shard_43.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.4.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "a29729f01de340dc23214e21eeff4c25" }, { "dataPath": "params_shard_44.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.3.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.3.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.3.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.4.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.4.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "f57b91218572f570f27389c7178af27e" }, { "dataPath": "params_shard_45.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.4.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "99d172b657d72b45a48a6082a69d8e91" }, { "dataPath": "params_shard_46.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.5.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "75d9c12a82e8f1c841724b16060277ec" }, { "dataPath": "params_shard_47.bin", "format": "raw-shard", "nbytes": 28682240, "records": [ { "name": "transformer.h.4.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.4.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.4.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.4.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.4.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 } ], "md5sum": "031af5a21e5f050925aad63a14af44be" }, { "dataPath": "params_shard_48.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.5.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "4fa8addb25101155d423d66f824edc6b" }, { "dataPath": "params_shard_49.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.10.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "c025a3579ac2eb2790479f983783a828" }, { "dataPath": "params_shard_50.bin", "format": "raw-shard", "nbytes": 28682240, "records": [ { "name": "transformer.h.5.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.5.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11468800 }, { "name": "transformer.h.5.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24576000 }, { "name": "transformer.h.5.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26214400 }, { "name": "transformer.h.10.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28672000 } ], "md5sum": "940aeaf04bb6a27d849ee1cf4eed60b2" }, { "dataPath": "params_shard_51.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.10.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "209785f8614fa2571d6134a1a272fda6" }, { "dataPath": "params_shard_52.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.10.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "0964082071ce303e0940d9203c73b8fa" }, { "dataPath": "params_shard_53.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.10.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.10.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.10.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.10.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.10.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "b74201a85cf9ac163cd596f518c6c753" }, { "dataPath": "params_shard_54.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.11.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "b48d3c57e510c84d3cf91f2d4d5470b3" }, { "dataPath": "params_shard_55.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.11.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "7797fad9d062edff5fa8e574a163912a" }, { "dataPath": "params_shard_56.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.10.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.11.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.11.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.11.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.11.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.11.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "b0e1818760373985606b625100a081f5" }, { "dataPath": "params_shard_57.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.12.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "9825b5c2b59287709d7f98abe474acc8" }, { "dataPath": "params_shard_58.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.12.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "dafabfb8ef2cae187eef1cff2be49966" }, { "dataPath": "params_shard_59.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.11.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.11.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.11.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.12.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.12.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "f2060378d2f4fbddbbfad888902f64c2" }, { "dataPath": "params_shard_60.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.12.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "33fb23827ac7aeb0f25dda44a7c85569" }, { "dataPath": "params_shard_61.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.5.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "61f866fbc422939b58327f0d5ec37dc4" }, { "dataPath": "params_shard_62.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.12.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.12.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.12.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.12.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.12.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.5.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "cc2f57ea67e46c06ad46324759d0b07c" }, { "dataPath": "params_shard_63.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.6.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "e2679bc2bc34222f03177d92af065d0a" }, { "dataPath": "params_shard_64.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.6.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "0880571476f03ad5e61a42e1eb208ba0" }, { "dataPath": "params_shard_65.bin", "format": "raw-shard", "nbytes": 22968320, "records": [ { "name": "transformer.h.5.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.5.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 5734400 }, { "name": "transformer.h.6.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 5744640 }, { "name": "transformer.h.6.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 5754880 }, { "name": "transformer.h.6.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 11489280 }, { "name": "transformer.h.6.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 22958080 } ], "md5sum": "b193d3b110fac4a722c1ff16ba1734ae" }, { "dataPath": "params_shard_66.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.6.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "fd5098b035acc5135a6f25aa10bc74e5" }, { "dataPath": "params_shard_67.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.7.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "7f1cf0bf87f81d0e6433c80dad24532c" }, { "dataPath": "params_shard_68.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.7.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "2f33c1e03a01c184c6bcda8a76ab9f2c" }, { "dataPath": "params_shard_69.bin", "format": "raw-shard", "nbytes": 22947840, "records": [ { "name": "transformer.h.6.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.6.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.6.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 14745600 }, { "name": "transformer.h.7.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.7.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 17213440 } ], "md5sum": "3c269a595ea271acf9691eab8d05bfa4" }, { "dataPath": "params_shard_70.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.7.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "3f580e04f5d79c51c0e922575edfa917" }, { "dataPath": "params_shard_71.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.8.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "d2cb6d43bf94b62b346a64d36d12e002" }, { "dataPath": "params_shard_72.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.7.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.7.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.7.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.7.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.7.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.8.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "2f165ecc307e55a15824fbaf0656f99f" }, { "dataPath": "params_shard_73.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.8.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "487a9e7e08cc9f606b4598019197a4a2" }, { "dataPath": "params_shard_74.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.8.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "3e84c5efe36c4c46d9201e5368947899" }, { "dataPath": "params_shard_75.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.8.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.8.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.8.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.8.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.8.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "b7f6d41902cb650b5d9bd29477a44561" }, { "dataPath": "params_shard_76.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.9.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "243b3ff7355ab50efd73423dedf48a0b" }, { "dataPath": "params_shard_77.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.9.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "1d4c9fc7f178935cd4097840654c8719" }, { "dataPath": "params_shard_78.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.8.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.9.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.9.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.9.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.9.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.9.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "ff5bcf574fc37a93822d18c4d6ab02ba" }, { "dataPath": "params_shard_79.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.13.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "d7e30542613ba57cf20cc404d122726a" }, { "dataPath": "params_shard_80.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.13.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "c7116cb6aaa82aa79cd03e3c5de4416a" }, { "dataPath": "params_shard_81.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.9.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.9.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.9.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.13.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.13.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "d3d1e5e27e923736b9010b5cdbd6e536" }, { "dataPath": "params_shard_82.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.13.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "5a89aad6bec36930a15608260cd76f13" }, { "dataPath": "params_shard_83.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.14.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "6144441caa20b9498398486e26223334" }, { "dataPath": "params_shard_84.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.13.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.13.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.13.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.13.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.13.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.14.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "960c5605023ae1d38e6307cba1e088db" }, { "dataPath": "params_shard_85.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.14.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "57c63dedc9b9765357eb267c5a7028b1" }, { "dataPath": "params_shard_86.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.14.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "95e8517ed7b7a38e1ec1e4c599adcaf5" }, { "dataPath": "params_shard_87.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.14.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.14.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.14.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.14.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.14.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "6c865c70e4975b6ff6fdbb38dba5cc41" }, { "dataPath": "params_shard_88.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.15.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "daae44b5776e57566937f84f7c01f028" }, { "dataPath": "params_shard_89.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.15.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "66e1b8752c2a3269d22764b98ffcd04c" }, { "dataPath": "params_shard_90.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.14.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.15.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.15.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.15.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.15.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.15.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "3965bf962b6f939d20a7d0a6d5efd18f" }, { "dataPath": "params_shard_91.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.16.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "2fdc9b979b55bc9269eace4fdff21fc1" }, { "dataPath": "params_shard_92.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.16.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "171df9861dfd8417de71c5d590d839f5" }, { "dataPath": "params_shard_93.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.15.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.15.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.15.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.16.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.16.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "c9f6b5a13af9be6c6ee51203e49f0b85" }, { "dataPath": "params_shard_94.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.16.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "dbc51484d57fbaa73a2cfa2ccdb00ba8" }, { "dataPath": "params_shard_95.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.17.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "4a32aa13c38342e37bc18bf18f0ed23c" }, { "dataPath": "params_shard_96.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.16.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.16.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.16.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.16.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.16.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.17.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "c72b2019124998dcdd312a4de3831923" }, { "dataPath": "params_shard_97.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.17.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "813d35a7ed2e8633e675ac6df32c716c" }, { "dataPath": "params_shard_98.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.17.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "691db977abf3a76b01945f0e9d6bc8c8" }, { "dataPath": "params_shard_99.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.17.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.17.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.17.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.17.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.17.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "214b97adff57b5ae8ece28a85438a38e" }, { "dataPath": "params_shard_100.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.18.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "fc3098fec5bcd846daf01644862572f5" }, { "dataPath": "params_shard_101.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.18.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "9ab002096820d865051d6670c0351b45" }, { "dataPath": "params_shard_102.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.17.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.18.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.18.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.18.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.18.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.18.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "48c6888986e9b58d8f3e5cb09cb5b417" }, { "dataPath": "params_shard_103.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.19.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "adaea0a6f271e7b40e00f3478e0904ff" }, { "dataPath": "params_shard_104.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.19.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "0f01e094f68f09ab1078939d4a6808b0" }, { "dataPath": "params_shard_105.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.18.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.18.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.18.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.19.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.19.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "03ad1d8737d8a50ca9888920986afda5" }, { "dataPath": "params_shard_106.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.19.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "512075f84683df7acea3586d5ecaa13d" }, { "dataPath": "params_shard_107.bin", "format": "raw-shard", "nbytes": 28682240, "records": [ { "name": "transformer.h.19.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.19.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.19.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.19.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.19.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 } ], "md5sum": "8398334763aaf3d8bb9be7503beb9461" }, { "dataPath": "params_shard_108.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.20.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "10bbf4798c2cc5c05d2d2f8a25db80f9" }, { "dataPath": "params_shard_109.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.20.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "37871bab0fa1554b61282b81812bcdef" }, { "dataPath": "params_shard_110.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.20.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "80ced229950e0be99735d64928ec27a1" }, { "dataPath": "params_shard_111.bin", "format": "raw-shard", "nbytes": 22947840, "records": [ { "name": "transformer.h.20.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.20.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.20.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 14745600 }, { "name": "transformer.h.20.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.20.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 17213440 } ], "md5sum": "dd7f482485823b6e7d37beb9c90e32a6" }, { "dataPath": "params_shard_112.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.21.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "746d861c1d1d3964c98f75b8d5e6efec" }, { "dataPath": "params_shard_113.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.21.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "78a24833bb8d96144614c83f0ea9776b" }, { "dataPath": "params_shard_114.bin", "format": "raw-shard", "nbytes": 28702720, "records": [ { "name": "transformer.h.20.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.20.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.21.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11479040 }, { "name": "transformer.h.21.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 11489280 }, { "name": "transformer.h.21.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 17223680 }, { "name": "transformer.h.21.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28692480 } ], "md5sum": "b62e0257656dfafd0a0e1454369df3f6" }, { "dataPath": "params_shard_115.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.21.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "70f3e8f6061c2e51dfa525249e6787f4" }, { "dataPath": "params_shard_116.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.22.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "03ac6340434623a0dda18831572b866e" }, { "dataPath": "params_shard_117.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.22.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "9fe755f8173ecac2ffb340c0e3758dc1" }, { "dataPath": "params_shard_118.bin", "format": "raw-shard", "nbytes": 22947840, "records": [ { "name": "transformer.h.21.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.21.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.21.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 14745600 }, { "name": "transformer.h.22.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.22.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 17213440 } ], "md5sum": "db7c4e0b99039080b09ee8bdbec4813a" }, { "dataPath": "params_shard_119.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.22.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "dcf8b8e018e3e403572545ddd39f2168" }, { "dataPath": "params_shard_120.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.23.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "519ceb72a2c7c9ea0a9c78b37578080d" }, { "dataPath": "params_shard_121.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.22.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.22.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.22.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.22.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.22.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.23.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "67dbe4f70bf3f30329025cfec7ba102b" }, { "dataPath": "params_shard_122.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.23.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "1f7c15a98b6355423548910d91505c57" }, { "dataPath": "params_shard_123.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.23.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "18518495e479f5aa21886a69cbada440" }, { "dataPath": "params_shard_124.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.23.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.23.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.23.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.23.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.23.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "4d6fc3376a51b4ba004b44526b749a93" }, { "dataPath": "params_shard_125.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.24.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "48bc110da189a0adff3821f6c867423c" }, { "dataPath": "params_shard_126.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.24.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "80a95040df616f7f9dbd22d7f912d5a6" }, { "dataPath": "params_shard_127.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.23.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.24.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.24.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.24.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.24.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.24.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "a8b99533bba4d0be8635e8b8512fdb07" }, { "dataPath": "params_shard_128.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.25.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "5e2e5e6d0fec465c2af8bdeb11325b19" }, { "dataPath": "params_shard_129.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.25.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "b4a999b9cd2ca2be189a9983a508a7e6" }, { "dataPath": "params_shard_130.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.24.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.24.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.24.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.25.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.25.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "7007abe4a52de16e40f24a7f97591557" }, { "dataPath": "params_shard_131.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.25.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "d7f4fa0acde36e495a9f2b34f76b97ba" }, { "dataPath": "params_shard_132.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.26.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "1daac0cac9097b53e61da0ac58537260" }, { "dataPath": "params_shard_133.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.25.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.25.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.25.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.25.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.25.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.26.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "9d22e77770b44976e4a8c0d343831701" }, { "dataPath": "params_shard_134.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.26.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "39ff1fd7809302a906153b8bb24fbf15" }, { "dataPath": "params_shard_135.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.26.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "e95ae48b7344a60785560b11c392574a" }, { "dataPath": "params_shard_136.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.26.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.26.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.26.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.26.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.26.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "8e2b0595f72d67f89b517c011b9c8424" }, { "dataPath": "params_shard_137.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.27.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "1a511de64c5bcdde1cbe5e84944b2b30" }, { "dataPath": "params_shard_138.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.27.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "358d01144cd09f0d7c7237ea6a91db34" }, { "dataPath": "params_shard_139.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.27.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "8fcc2530afe416e1879a481086c01d85" }, { "dataPath": "params_shard_140.bin", "format": "raw-shard", "nbytes": 25405440, "records": [ { "name": "transformer.h.26.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.27.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 2457600 }, { "name": "transformer.h.27.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 15564800 }, { "name": "transformer.h.27.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 17203200 }, { "name": "transformer.h.27.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19660800 }, { "name": "transformer.h.27.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 19671040 } ], "md5sum": "e5421b5e4d843355a46a2dd9428556c5" }, { "dataPath": "params_shard_141.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.28.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "4fe8dcd1f718be8142a72e7c71135d6e" }, { "dataPath": "params_shard_142.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.28.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "4fa91b61adfa4a7718bc613edfdcc0ff" }, { "dataPath": "params_shard_143.bin", "format": "raw-shard", "nbytes": 28702720, "records": [ { "name": "transformer.h.27.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.27.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.28.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11479040 }, { "name": "transformer.h.28.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 11489280 }, { "name": "transformer.h.28.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 17223680 }, { "name": "transformer.h.28.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28692480 } ], "md5sum": "e5778d78b462509218de42980f76e58e" }, { "dataPath": "params_shard_144.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.28.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "22d586e247f801145eb10aa53f2e89e4" }, { "dataPath": "params_shard_145.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.29.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "a1a201413096e1c4f7e1237cdebb3bbb" }, { "dataPath": "params_shard_146.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.29.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "16b8c6625e9eb28b8d3308197a0211a2" }, { "dataPath": "params_shard_147.bin", "format": "raw-shard", "nbytes": 22947840, "records": [ { "name": "transformer.h.28.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.28.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.28.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 14745600 }, { "name": "transformer.h.29.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.29.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 17213440 } ], "md5sum": "43bcc6fdc2f7730093cd430355f8ae5b" }, { "dataPath": "params_shard_148.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.29.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "a6fa396e516c6b010469bd14c5d701b1" }, { "dataPath": "params_shard_149.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.30.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "a01e64f49bc5297e256a51032504a995" }, { "dataPath": "params_shard_150.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.29.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.29.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.29.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.29.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.29.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.30.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "cee6b749750f3fca3ad7f675c877ffe8" }, { "dataPath": "params_shard_151.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.30.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "d5b1538f327f6565a96a646e2f250841" }, { "dataPath": "params_shard_152.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.30.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "5eb4e4c9cc01710f27655cf52e34d67e" }, { "dataPath": "params_shard_153.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.30.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.30.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.30.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.30.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.30.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "9c70c0936b5e655dda68023d2d7f078e" }, { "dataPath": "params_shard_154.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.31.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "8644b0bc3c9313a4e9e24b67ea009a3a" }, { "dataPath": "params_shard_155.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.31.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "050206b2ce4c66cceaf719ce2eaf0e6c" }, { "dataPath": "params_shard_156.bin", "format": "raw-shard", "nbytes": 32788480, "records": [ { "name": "transformer.h.30.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.31.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 2457600 }, { "name": "transformer.h.31.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 2467840 }, { "name": "transformer.h.31.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 8202240 }, { "name": "transformer.h.31.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 19671040 }, { "name": "transformer.h.31.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19681280 } ], "md5sum": "2fcee12d653b04ced742de171553d011" }, { "dataPath": "params_shard_157.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.32.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "95afbb7f32c80ded9f49e89b97765d89" }, { "dataPath": "params_shard_158.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.32.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "d345880adf672671329df099389d17c9" }, { "dataPath": "params_shard_159.bin", "format": "raw-shard", "nbytes": 29501440, "records": [ { "name": "transformer.h.31.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.31.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 1638400 }, { "name": "transformer.h.31.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 21299200 }, { "name": "transformer.h.32.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 23756800 }, { "name": "transformer.h.32.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 23767040 } ], "md5sum": "aa351e3bf2806156a88218d383a0d4cd" }, { "dataPath": "params_shard_160.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.32.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "844b5db12492eebf4d3592b153a4a7b1" }, { "dataPath": "params_shard_161.bin", "format": "raw-shard", "nbytes": 45875200, "records": [ { "name": "transformer.h.33.mlp.down_proj.q_weight", "shape": [ 5120, 2240 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 45875200, "byteOffset": 0 } ], "md5sum": "fdd8c94865e922629afca686b3e866cf" }, { "dataPath": "params_shard_162.bin", "format": "raw-shard", "nbytes": 28692480, "records": [ { "name": "transformer.h.32.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 0 }, { "name": "transformer.h.32.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 11468800 }, { "name": "transformer.h.32.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 11479040 }, { "name": "transformer.h.32.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 24586240 }, { "name": "transformer.h.32.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 26224640 }, { "name": "transformer.h.33.ln.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 28682240 } ], "md5sum": "de5d54d3cf4e10ad6ca7ca1fa11b77f0" }, { "dataPath": "params_shard_163.bin", "format": "raw-shard", "nbytes": 91750400, "records": [ { "name": "transformer.h.33.mlp.gate_up_proj.q_weight", "shape": [ 35840, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 91750400, "byteOffset": 0 } ], "md5sum": "e6aeca269415c75da8910ea8b6de0f86" }, { "dataPath": "params_shard_164.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.33.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "c4d49a818837757140eff6acaf98b667" }, { "dataPath": "params_shard_165.bin", "format": "raw-shard", "nbytes": 31959040, "records": [ { "name": "transformer.h.33.mlp.down_proj.q_scale", "shape": [ 5120, 560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5734400, "byteOffset": 0 }, { "name": "transformer.h.33.mlp.gate_up_proj.q_scale", "shape": [ 35840, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 11468800, "byteOffset": 5734400 }, { "name": "transformer.h.33.post_attention_layernorm.weight", "shape": [ 5120 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 10240, "byteOffset": 17203200 }, { "name": "transformer.h.33.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 17213440 }, { "name": "transformer.h.33.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 30320640 } ], "md5sum": "dbd91f633b7d5f58f79430d0dc1f4d26" }, { "dataPath": "params_shard_166.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.34.mixer.qkv_proj.q_weight", "shape": [ 7680, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 19660800, "byteOffset": 0 } ], "md5sum": "09dc4cf6a1ec031511c6e26a0f4e4681" }, { "dataPath": "params_shard_167.bin", "format": "raw-shard", "nbytes": 19660800, "records": [ { "name": "transformer.h.33.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 0 }, { "name": "transformer.h.34.mixer.out_proj.q_weight", "shape": [ 5120, 640 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 2457600 }, { "name": "transformer.h.34.mixer.out_proj.q_scale", "shape": [ 5120, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 15564800 }, { "name": "transformer.h.34.mixer.qkv_proj.q_scale", "shape": [ 7680, 160 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 2457600, "byteOffset": 17203200 } ], "md5sum": "1625c47fbda0f8a57b82a06fe9236af8" } ] }