diff --git "a/ndarray-cache.json" "b/ndarray-cache.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache.json" @@ -0,0 +1,11583 @@ +{ + "metadata": { + "ParamSize": 805, + "ParamBytes": 36381229568.0, + "BitsPerParam": 3.4891903342423802 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 525344768, + "records": [ + { + "name": "lm_head.q_weight", + "shape": [ + 128258, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 525344768, + "byteOffset": 0 + } + ], + "md5sum": "66dc1876fac8c95dc8fdd613e7d03662" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.78.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "316534b798d5b3d3be87417d5d96ae95" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.78.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "2ef693be7ee47a3b57f2831bcbaae1b5" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.79.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "cd2bc5b51f60ca4919c1ea3d39741ee2" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.79.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "0d967b9380c2686c32401448006dbf04" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 31146240, + "records": [ + { + "name": "lm_head.q_scale", + "shape": [ + 128258, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16417024, + "byteOffset": 0 + }, + { + "name": "model.layers.78.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 16417024 + }, + { + "name": "model.layers.78.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 16433408 + }, + { + "name": "model.layers.78.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 20103424 + }, + { + "name": "model.layers.78.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 27443456 + }, + { + "name": "model.layers.79.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 27459840 + }, + { + "name": "model.layers.79.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27476224 + } + ], + "md5sum": "34efe8e4d3a88f64eee2596c1cf501d7" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.79.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "8939fa981ecc0c1a14f56a8516318707" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.79.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "ca43026fe3a5cb8a0c379c38082aaf8e" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 525344768, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 128258, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 525344768, + "byteOffset": 0 + } + ], + "md5sum": "5638292ae0a24e47f369af28372d8a03" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "11879e9d050b12ec6f0d54da37479648" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "7beff83393e66099f91d631959b104ce" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 29835520, + "records": [ + { + "name": "model.layers.79.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.79.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.79.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.79.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.norm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 128258, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16417024, + "byteOffset": 9732096 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26149120 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 26165504 + } + ], + "md5sum": "a85cf2c3a495e5d6bb6e22979cfb1a0a" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.0.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "dd98fed988ac3dd8351cdfbf527af713" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "9a27773b6d5777d26bf8549c703db5aa" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "4839008d2bf9f6ab39a49a89e32bb5d4" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.1.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "611263b9c805dfbb7103f43bea629714" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "4b20edfa7f3f40dce4978e16b47e4ca6" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "34be46e8c9217aafde30226f55359d2d" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "c4d2b8a23575bdf0e2be054d86777a50" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "2633d1f0743a5de16281c492b3ae05f5" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 9715712 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 17055744 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 18366464 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19415040 + }, + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 19431424 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "28aaf4c5bbb7972492f5d95a641b6f71" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.2.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "af58f75984a713bb980ef0dbce8bb51c" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "60c3b22cc9999559cba1de7afdcf11fe" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "4759ae6eee5c1b9c200d3a29dfe5dda4" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "01dfb99a48c7f4c2a40f1df3f80fbfa2" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.3.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "18757c12edc6b778624be9ce40661db7" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "34e7f2899ac5b55bc96b9868f3e50350" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.4.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "6fef6e9a0a8905766f99fa8d4835d24c" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "12b362e0181ac9de3247c49eb9ff2a70" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "914478bac4ee9ea93b5ac8c99d03dadf" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "aa6009a6e934229257544d7d24fecf08" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 29163520, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 24428544 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 25493504 + } + ], + "md5sum": "55f8244d42c5f3ff48ef60bea3c27b1f" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.10.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "966da50a795e511c7a67ca3b1d860d07" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "f06bf977704a35420c5580da5523073f" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "e9aa927b58d99230b4bd3719640ff254" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "5607165dcc314fad710f0528efdd537f" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.11.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "710bb1e4b50e646e9683eb336a6518b8" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "e09e8fec2e917fe0c131ee2733111873" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "f296490ccf682507c4a2cb4a0ec26986" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.12.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "02f456681a1e57c5e3ee71147636e426" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c267f30607a86ec72c6b8ed9f03ef484" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "8415d96ccc83cc64fe1cb232dd64ea66" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 32833536, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 23117824 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 30457856 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 31768576 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 32817152 + } + ], + "md5sum": "0c7af87941c3b578754c33992e3039a3" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "498477964a2b5d2a86fefaf8b3585870" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "c3b974c1c3a32faf3c25999cbd8b9a99" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "bd28f00d55c72a89061888ec794a0dfb" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.13.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "d5d9a58ea5670c2d698988915ca52429" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "4d5f74fba9cb01f173cd1b30c8ae6d56" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "f281136674e030bfde72886783218d64" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "029bc18c24df44dc75686cdaec0430c7" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 28147712, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7356416 + }, + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 7372800 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11042816 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11059200 + }, + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 11075584 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 14745600 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22085632 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 22102016 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 23412736 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24461312 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 24477696 + } + ], + "md5sum": "0fb32cf26587111a007390832eee1b84" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.14.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "163e6e0cc2f856c390af4c65dfd84c50" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "b10f63cf5290beec6d7def21038f3728" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.15.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "681850b2dec664c254b775f1b71b60c6" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "b4b28de6c6e49fce4e3d41475c72168a" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "b927ba0b0d681b23cf959dbd0373e4fa" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "4234b9379714f1254b8a37ca15879e83" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "97ab7a3c423f3ba68458cb879a494185" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "b7f44aad83d6268ebab3345c3ebba280" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 9715712 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 11026432 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 12075008 + }, + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 12091392 + }, + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 15761408 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "1bd672f284d0ffed208bedcee38ac2d4" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.16.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "eaea28c72c299938be516cea681fad1e" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "24c08573e377ec521dd52c40faa5d211" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "008fe7c69e978cde48c6b9340dd7645e" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "b8f4693b2ea4d9893754049f71a3f18d" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.17.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "e399cb06e6f3913b330de94ca9e35dd2" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "29b11d3f85a1060a1187de201ae56caa" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.18.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "c1850f2836a161798a1bdc2ea23b9af8" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "6a7f576c96ee5a93b884093167b1ddbf" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "85f0941a4d0264de0fe69a682ab0b331" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 28114944, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24428544 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 24444928 + } + ], + "md5sum": "61e7509888611989e3dde8cfc30ad54f" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "ebb65c1d2a62bf7dcdd710a9b0374d37" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "4fd61337ce18bee5a2d3ab485a9e745f" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a0fc530369a1e637a9da4f1b7034d182" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.19.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "81558bcee7a0e008f31299139a05cc14" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "b7798e84eef18a2588eb4f79ab3a2227" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "2632e3ff022d96acd0193bd77f9fc659" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "e4727d2aaae443c22335f5c81ac37a2b" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.20.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "481c09e6150091c53ae7fe369750a916" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 32817152, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 7356416 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 8404992 + }, + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 8421376 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 12091392 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19431424 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 19447808 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 20758528 + }, + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 21807104 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 25477120 + } + ], + "md5sum": "71a06f859c9dcdffe2b77e8bbbfb23a9" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "bec6ffc8792b2be31f5ed33cec73ad3b" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "b6e33fc34ab63eddb211bcc732bd9b6b" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "0d8b62e3b4efec2b2c505b52c5a35393" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.21.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "7d3e037493d53b31c5cfd8c7c83c2218" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "60f7af9651081d7bfc2afcf32bc52dd2" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "e570b96985655c956ffe247a93126273" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "2a531b2d4a0fff022200d6536dc1a7c6" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.22.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "0a003a95813e32638e437246324afabc" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "4393c12cf06b0bea81f641b3a6fdef1a" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "c031d012d075df27d54fd394e8cdf8b6" + }, + { + "dataPath": "params_shard_89.bin", + "format": "raw-shard", + "nbytes": 29196288, + "records": [ + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 0 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 1310720 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2375680 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2392064 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2408448 + }, + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6078464 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13418496 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 13434880 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 14745600 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15794176 + }, + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 15810560 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 19480576 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26820608 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 26836992 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 28147712 + } + ], + "md5sum": "a2ec94425363a67510b4abd34d13928c" + }, + { + "dataPath": "params_shard_90.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.23.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "b9cf4242164343a53ea82ef58b804407" + }, + { + "dataPath": "params_shard_91.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "2717c7813a4693f730485356f8fb1e60" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "0a663ec0c3ff0c8e1b3dccfada875e7a" + }, + { + "dataPath": "params_shard_93.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "5c54a564a2c27de8f07fda14c16f4185" + }, + { + "dataPath": "params_shard_94.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "0830f7bb0818392382f5fa69a98ec314" + }, + { + "dataPath": "params_shard_95.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.24.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "0c33aed06a35579432af8a97025fcadb" + }, + { + "dataPath": "params_shard_96.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "0256d76138b1743d2dd371912a73f7f6" + }, + { + "dataPath": "params_shard_97.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "14cd76bd263f3ad31e19f16cd9099dcd" + }, + { + "dataPath": "params_shard_98.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "1b1e57516df66bf966fbd53f7d4e81b8" + }, + { + "dataPath": "params_shard_99.bin", + "format": "raw-shard", + "nbytes": 30490624, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9715712 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13402112 + }, + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 13418496 + }, + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 17088512 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24428544 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 24444928 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 25755648 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26804224 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 26820608 + } + ], + "md5sum": "a4dc042109cfec1fe1d64a4fcd4852f8" + }, + { + "dataPath": "params_shard_100.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.25.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "a12f6976571630ac814e369c3d93d11e" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "ffcc41e957b177ad6c102a42c2c2f02c" + }, + { + "dataPath": "params_shard_102.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.26.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "5f6d72e62dc053303991ecb2ced29b07" + }, + { + "dataPath": "params_shard_103.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "cfe1cbc9795e3b56418b7356aeb320d2" + }, + { + "dataPath": "params_shard_104.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a3b51598415593b8467a4ba514c50e88" + }, + { + "dataPath": "params_shard_105.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "db474d77e6c0868920835b62acbca454" + }, + { + "dataPath": "params_shard_106.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "f62ef97766b150235bcb552b2ac10567" + }, + { + "dataPath": "params_shard_107.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "596e75ff9312dddc81ed64bda241a6b3" + }, + { + "dataPath": "params_shard_108.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 9715712 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 11026432 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 12075008 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 12091392 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 15761408 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "219fa2e316b1d22873a363daefd2907a" + }, + { + "dataPath": "params_shard_109.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.27.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "c0c10b5509d828130d36672de7561886" + }, + { + "dataPath": "params_shard_110.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "a5f9d66481bc3571f0f19d339567b2bb" + }, + { + "dataPath": "params_shard_111.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "bceccb8d88b0ce1fa5077bc38d5ab069" + }, + { + "dataPath": "params_shard_112.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "4991b4be0cc55f790c6ab4698b53e33c" + }, + { + "dataPath": "params_shard_113.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.28.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "8fac0e506152d6e03b3f8d6e37233636" + }, + { + "dataPath": "params_shard_114.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "fe7ea9ec0c03ecbe2a58654ef06ac429" + }, + { + "dataPath": "params_shard_115.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.29.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "1bb2bf97fa014cedf507147ba8b6b054" + }, + { + "dataPath": "params_shard_116.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a9d65785c1dc81ed02896aa96cc85152" + }, + { + "dataPath": "params_shard_117.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "fe7140df56d6815928b3ffde0deb4555" + }, + { + "dataPath": "params_shard_118.bin", + "format": "raw-shard", + "nbytes": 28114944, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24428544 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 24444928 + } + ], + "md5sum": "953e457515f6ee622c89b193121cbd88" + }, + { + "dataPath": "params_shard_119.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "d26ade2d30fbf06d18d4d11d5ec8e32e" + }, + { + "dataPath": "params_shard_120.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "99a627c92bc4a803618dddd8e6339c96" + }, + { + "dataPath": "params_shard_121.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "29b7d69bcc0f8601841e8ba0aaa1afe6" + }, + { + "dataPath": "params_shard_122.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.30.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "e913a94fba758817db7248da263b7eb0" + }, + { + "dataPath": "params_shard_123.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "9b740bfd881ab27ed7127988a6875ea9" + }, + { + "dataPath": "params_shard_124.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "bd8213e3bca1676f75f909c60752ca46" + }, + { + "dataPath": "params_shard_125.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "cfcbb9c2f05df3714f7ced4ca9d7dcf2" + }, + { + "dataPath": "params_shard_126.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.31.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "909acd1104248371a46df3905e1e8e32" + }, + { + "dataPath": "params_shard_127.bin", + "format": "raw-shard", + "nbytes": 32817152, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 7356416 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 8404992 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 8421376 + }, + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 12091392 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19431424 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 19447808 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 20758528 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 21807104 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 25477120 + } + ], + "md5sum": "b6557eb39e3051189b9a094bf70826b2" + }, + { + "dataPath": "params_shard_128.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "b6efde09b225fb1183b934d52c667f1f" + }, + { + "dataPath": "params_shard_129.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.32.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "85a6fcfa58e52f370289c4f5764670b1" + }, + { + "dataPath": "params_shard_130.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "f368698ae2dbac3b8eee395ab6255fd5" + }, + { + "dataPath": "params_shard_131.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.32.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "6ca066a40968c58283d968067838a797" + }, + { + "dataPath": "params_shard_132.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.32.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "14770b6f9907a04aa67c01215d341d02" + }, + { + "dataPath": "params_shard_133.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "239be40a1e00055d672e1434143bdaf7" + }, + { + "dataPath": "params_shard_134.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "4c607f754f8a7172fd7116a98a370dcf" + }, + { + "dataPath": "params_shard_135.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.33.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "c2f22e174feb17caf571ff5c795c7ab5" + }, + { + "dataPath": "params_shard_136.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.33.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "011fa10f08032aef4d5de0fdc8704bd2" + }, + { + "dataPath": "params_shard_137.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "0cc95474be25775c2ceecadb69fdd185" + }, + { + "dataPath": "params_shard_138.bin", + "format": "raw-shard", + "nbytes": 29196288, + "records": [ + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 0 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 1310720 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2375680 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2392064 + }, + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2408448 + }, + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6078464 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13418496 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 13434880 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 14745600 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15794176 + }, + { + "name": "model.layers.33.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 15810560 + }, + { + "name": "model.layers.33.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 19480576 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26820608 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 26836992 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 28147712 + } + ], + "md5sum": "ca393f6c7cdc225d2b7b2076b13c75f0" + }, + { + "dataPath": "params_shard_139.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.34.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "8222d8438d1da23b1dbafd43ee52d172" + }, + { + "dataPath": "params_shard_140.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.34.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "fcb3d07d78a1dfcbd21961d0d3ba5ba7" + }, + { + "dataPath": "params_shard_141.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.34.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "59edd0136bd2e9ada69bc166453f0644" + }, + { + "dataPath": "params_shard_142.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "d8aede90391c0f300db4b1cf9c1b2c1c" + }, + { + "dataPath": "params_shard_143.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.35.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "1e5a1a91d38f3b81bceba808a276143d" + }, + { + "dataPath": "params_shard_144.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.35.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "381a908226812d90e0114d368d6901f8" + }, + { + "dataPath": "params_shard_145.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.35.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "7347af2502509695eba41dcda3daab89" + }, + { + "dataPath": "params_shard_146.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.36.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "914feae43e5f3ee27ddff66d228e4004" + }, + { + "dataPath": "params_shard_147.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "f7550997f3e91de8ba41f00f48579ac0" + }, + { + "dataPath": "params_shard_148.bin", + "format": "raw-shard", + "nbytes": 30490624, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.34.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9715712 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13402112 + }, + { + "name": "model.layers.35.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 13418496 + }, + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 17088512 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24428544 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 24444928 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 25755648 + }, + { + "name": "model.layers.36.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26804224 + }, + { + "name": "model.layers.36.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 26820608 + } + ], + "md5sum": "4257418863ec7c4afa9713e161f2a817" + }, + { + "dataPath": "params_shard_149.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.36.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "dfa60a22b7d162ddbeee37fc0b9043d7" + }, + { + "dataPath": "params_shard_150.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.36.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "6b6ffa0b1975640a877e5cc57f8ef70d" + }, + { + "dataPath": "params_shard_151.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.37.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "bf960efa868815e7713596f04d9f0ceb" + }, + { + "dataPath": "params_shard_152.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.37.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "60fb94b77574e513df0a110a6a48f91f" + }, + { + "dataPath": "params_shard_153.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a94c37fb2084b47358e5b07df5b26277" + }, + { + "dataPath": "params_shard_154.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.37.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "de3e3f2dabc7e643f38dfa36743bace1" + }, + { + "dataPath": "params_shard_155.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.38.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "06417fc20e7b87cc400f8fa91d57ec05" + }, + { + "dataPath": "params_shard_156.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "b4eb5f6ad6dec26d20a740dc06ef0e88" + }, + { + "dataPath": "params_shard_157.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.36.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 9715712 + }, + { + "name": "model.layers.37.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 11026432 + }, + { + "name": "model.layers.37.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 12075008 + }, + { + "name": "model.layers.37.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 12091392 + }, + { + "name": "model.layers.37.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 15761408 + }, + { + "name": "model.layers.37.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.38.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.38.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "38b9a0c18144bb368b432987cb544d29" + }, + { + "dataPath": "params_shard_158.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.38.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "1131811f13f0972ab2b93aa9cc769db0" + }, + { + "dataPath": "params_shard_159.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.38.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c0c435e0d50a02ec11db200b6842bc9e" + }, + { + "dataPath": "params_shard_160.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "2c7d5db8076d82c4dcef366c0e8a9e24" + }, + { + "dataPath": "params_shard_161.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.39.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "01e4aba200e8f78c7f834437c4bce14f" + }, + { + "dataPath": "params_shard_162.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.39.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "0010cb9a7933ddf482f337d1ec023ebb" + }, + { + "dataPath": "params_shard_163.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.39.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "8c17112891103d0210ff7f092032c6fa" + }, + { + "dataPath": "params_shard_164.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.40.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "d60673516414dedb59289f76537fac2d" + }, + { + "dataPath": "params_shard_165.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "9c9d4419fc3f4c90c79e637e7d08f4f7" + }, + { + "dataPath": "params_shard_166.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "abcf0f1b02e3f382f7e2d507dd0665f4" + }, + { + "dataPath": "params_shard_167.bin", + "format": "raw-shard", + "nbytes": 28114944, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.38.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.39.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.39.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.39.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.39.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.39.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24428544 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 24444928 + } + ], + "md5sum": "57a1dd5438dfa6c8ae45a1542b46f30c" + }, + { + "dataPath": "params_shard_168.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "f5933b9929d41434141e035ee3528519" + }, + { + "dataPath": "params_shard_169.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "05dd2304cb9d81fcb96c1e5514fb153b" + }, + { + "dataPath": "params_shard_170.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.5.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "ecffe9ee5d72a5d7737eacffd2e260a8" + }, + { + "dataPath": "params_shard_171.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "0fb6e6bae047df9760dbb6fb3876b442" + }, + { + "dataPath": "params_shard_172.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "ffe0a81d93f853d2f1f7c653259c4961" + }, + { + "dataPath": "params_shard_173.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "174f825748f0ca84e8e055d65fcdba17" + }, + { + "dataPath": "params_shard_174.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.6.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "7aa87b71191aa73ceadf33e340a37ea4" + }, + { + "dataPath": "params_shard_175.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "2bedaa83f5846d88bdc95807b3851979" + }, + { + "dataPath": "params_shard_176.bin", + "format": "raw-shard", + "nbytes": 33112064, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7356416 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 7372800 + }, + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 11042816 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 18382848 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 18399232 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 19709952 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 20774912 + }, + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 24444928 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31784960 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 31801344 + } + ], + "md5sum": "3425acdf219231b7900291701586998e" + }, + { + "dataPath": "params_shard_177.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.7.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "4bc3517ad56a1d9e2799d5de034fdc51" + }, + { + "dataPath": "params_shard_178.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "4dbc98d4f625580092feacbd95ca87a3" + }, + { + "dataPath": "params_shard_179.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.40.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "26cc16a327e2898b89f2a0af5c3581aa" + }, + { + "dataPath": "params_shard_180.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.40.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "3dc35a82ef9e67efba3852d5f088d40d" + }, + { + "dataPath": "params_shard_181.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.41.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "7dd178da03fb3ac87dca5a3cbf2f36c3" + }, + { + "dataPath": "params_shard_182.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "99109df0d7be65f7c2585311570169d4" + }, + { + "dataPath": "params_shard_183.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.41.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "813be2df8038e792989e894d8a1c23a9" + }, + { + "dataPath": "params_shard_184.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.41.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "a8f671e6e1dbc63998e5c83cf0e13f28" + }, + { + "dataPath": "params_shard_185.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.42.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "ce80ea21b72e96b66ec7f2640fa9467f" + }, + { + "dataPath": "params_shard_186.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.42.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "eb8a501b3579b4584669ab88f6912d91" + }, + { + "dataPath": "params_shard_187.bin", + "format": "raw-shard", + "nbytes": 31522816, + "records": [ + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 0 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 1048576 + }, + { + "name": "model.layers.40.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.40.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2375680 + }, + { + "name": "model.layers.40.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6045696 + }, + { + "name": "model.layers.40.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.40.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 13402112 + }, + { + "name": "model.layers.41.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 14450688 + }, + { + "name": "model.layers.41.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 14467072 + }, + { + "name": "model.layers.41.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 18137088 + }, + { + "name": "model.layers.41.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 25493504 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 26804224 + }, + { + "name": "model.layers.42.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27852800 + } + ], + "md5sum": "be9ab42783603c2c9869177c8902a284" + }, + { + "dataPath": "params_shard_188.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.42.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "30d92f0d18d6a22837162e608063cd04" + }, + { + "dataPath": "params_shard_189.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.42.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c624d8f9c3058ddb826f418ff9e3a03e" + }, + { + "dataPath": "params_shard_190.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.43.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "3bc11a36a86744ef235fee870ad6ca4f" + }, + { + "dataPath": "params_shard_191.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.43.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "9f61fb97a16abb80a7feb8eb0ca37251" + }, + { + "dataPath": "params_shard_192.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.43.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "eb4f33f77deca4635d73d048ba5b1443" + }, + { + "dataPath": "params_shard_193.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.43.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "1bc81a5e45813cd92300f31ceaf51b98" + }, + { + "dataPath": "params_shard_194.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.44.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "e822f34213b569cc31fbd51a7a48f2bd" + }, + { + "dataPath": "params_shard_195.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.44.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a898b3a34d39773b783ad57ff18cbe36" + }, + { + "dataPath": "params_shard_196.bin", + "format": "raw-shard", + "nbytes": 26820608, + "records": [ + { + "name": "model.layers.42.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.42.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.42.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.42.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.42.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.43.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9732096 + }, + { + "name": "model.layers.43.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9748480 + }, + { + "name": "model.layers.43.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13418496 + }, + { + "name": "model.layers.43.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.43.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20774912 + }, + { + "name": "model.layers.43.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22085632 + }, + { + "name": "model.layers.44.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23134208 + }, + { + "name": "model.layers.44.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23150592 + } + ], + "md5sum": "97d369f8e1f5496ea2699c679d044b2e" + }, + { + "dataPath": "params_shard_197.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.44.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "32bad65ff435a511c1c980705fd2eb24" + }, + { + "dataPath": "params_shard_198.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.44.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "960d4f098f33eb81901b0b3323dd635b" + }, + { + "dataPath": "params_shard_199.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.45.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "21b35e2fcc696d46f4cb7c7bb661e32b" + }, + { + "dataPath": "params_shard_200.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.45.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "854a192a02daa22b63eea5d7c9bab022" + }, + { + "dataPath": "params_shard_201.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.45.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "bb71cfedb7434faca75e07740eb83956" + }, + { + "dataPath": "params_shard_202.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.45.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "c09792cb616fb83d123606a677e3a031" + }, + { + "dataPath": "params_shard_203.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.46.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "94638b8298fa440c4435f04eb8ce9c3c" + }, + { + "dataPath": "params_shard_204.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.46.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "3a1ad5acb4cb4051b5ecee3bd9a4b2cc" + }, + { + "dataPath": "params_shard_205.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.44.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.44.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.44.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.44.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.45.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 9715712 + }, + { + "name": "model.layers.45.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 17055744 + }, + { + "name": "model.layers.45.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 18366464 + }, + { + "name": "model.layers.45.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19415040 + }, + { + "name": "model.layers.45.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 19431424 + }, + { + "name": "model.layers.45.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.46.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.46.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "3fcaad8f8909736d7ebdc586fb9db2ad" + }, + { + "dataPath": "params_shard_206.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.46.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "67befa459f5ad856222d8ddf4a1f1cbc" + }, + { + "dataPath": "params_shard_207.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.46.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "133fb970d80da48ec9c2538b93c5b873" + }, + { + "dataPath": "params_shard_208.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.47.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "70fb2a20a6977413f5b36f6b73ceea30" + }, + { + "dataPath": "params_shard_209.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.47.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "e97323be48821636adc9b3c7733eedbb" + }, + { + "dataPath": "params_shard_210.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.47.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "2d439b5baa087a05df40f92b5804fe17" + }, + { + "dataPath": "params_shard_211.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.47.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c6d329930ed0f093bffcdf71f1efc391" + }, + { + "dataPath": "params_shard_212.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.48.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "90b97421a9f433d6d861632274f7bcb6" + }, + { + "dataPath": "params_shard_213.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.48.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "75b4a44e64446bfadd689d61af44619b" + }, + { + "dataPath": "params_shard_214.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.48.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "4223cc1960304cd7b5d94059c5c3eba9" + }, + { + "dataPath": "params_shard_215.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.48.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "0a9f2fc95ec492c26b10acd451357814" + }, + { + "dataPath": "params_shard_216.bin", + "format": "raw-shard", + "nbytes": 29163520, + "records": [ + { + "name": "model.layers.46.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.46.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.46.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.46.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.47.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.47.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.47.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.47.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.47.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.47.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.48.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.48.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 24428544 + }, + { + "name": "model.layers.48.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.48.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 25493504 + } + ], + "md5sum": "f0144de42afa1c168ef2fed3ac09af31" + }, + { + "dataPath": "params_shard_217.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.49.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "9a2e4ab30e0f5536eea225552a8117e3" + }, + { + "dataPath": "params_shard_218.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.49.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "5e6e018b499e10b6762179804211de61" + }, + { + "dataPath": "params_shard_219.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.49.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "eb1bf782724cd4267bbeafb6337f6e8f" + }, + { + "dataPath": "params_shard_220.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.49.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c76484cf062d235043cece5a69dca43a" + }, + { + "dataPath": "params_shard_221.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.50.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "f2048f0819d2b24c5e9e8390084e5890" + }, + { + "dataPath": "params_shard_222.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.50.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "050a9d5b2b1a68de165902b19a06412d" + }, + { + "dataPath": "params_shard_223.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.50.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "c043f32a085593c1b6605d629d349544" + }, + { + "dataPath": "params_shard_224.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.50.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "056c9f624c13a3a51f15d9817f082a82" + }, + { + "dataPath": "params_shard_225.bin", + "format": "raw-shard", + "nbytes": 33112064, + "records": [ + { + "name": "model.layers.48.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.48.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.49.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7356416 + }, + { + "name": "model.layers.49.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 7372800 + }, + { + "name": "model.layers.49.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 11042816 + }, + { + "name": "model.layers.49.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 18382848 + }, + { + "name": "model.layers.49.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 18399232 + }, + { + "name": "model.layers.49.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 19709952 + }, + { + "name": "model.layers.50.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.50.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 20774912 + }, + { + "name": "model.layers.50.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 24444928 + }, + { + "name": "model.layers.50.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31784960 + }, + { + "name": "model.layers.50.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 31801344 + } + ], + "md5sum": "45efbe95de9b6e8c6908c51d87f35871" + }, + { + "dataPath": "params_shard_226.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.51.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "a0023f297549ec509e0aac9f80b3f9b9" + }, + { + "dataPath": "params_shard_227.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.51.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "9205dda1e874c90c654c1bf6e40c4c62" + }, + { + "dataPath": "params_shard_228.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.51.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "3ea8b9567ddedde0b75bf57ca55c04ac" + }, + { + "dataPath": "params_shard_229.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.51.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "9357faa17496348a9e163d8d565d2d45" + }, + { + "dataPath": "params_shard_230.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.52.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "c901c0212bacaddbfeefa220ce4988a7" + }, + { + "dataPath": "params_shard_231.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.52.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "cf6e48f74749ee1b05b86026489d967c" + }, + { + "dataPath": "params_shard_232.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.52.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "22f56f3276fa3b828ea60cf68788c340" + }, + { + "dataPath": "params_shard_233.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.52.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "1a4bd8eae95591968163549390d59a09" + }, + { + "dataPath": "params_shard_234.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.53.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "6482a2bd45f33df20bd571676ca10d7c" + }, + { + "dataPath": "params_shard_235.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.53.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "f029480f618fda5a12aca9a9a6374f25" + }, + { + "dataPath": "params_shard_236.bin", + "format": "raw-shard", + "nbytes": 31522816, + "records": [ + { + "name": "model.layers.50.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 0 + }, + { + "name": "model.layers.51.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 1048576 + }, + { + "name": "model.layers.51.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.51.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2375680 + }, + { + "name": "model.layers.51.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6045696 + }, + { + "name": "model.layers.51.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.51.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 13402112 + }, + { + "name": "model.layers.52.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 14450688 + }, + { + "name": "model.layers.52.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 14467072 + }, + { + "name": "model.layers.52.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 18137088 + }, + { + "name": "model.layers.52.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.52.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 25493504 + }, + { + "name": "model.layers.52.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 26804224 + }, + { + "name": "model.layers.53.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27852800 + } + ], + "md5sum": "03ec43d9c83d05d059e662c00d41dbdf" + }, + { + "dataPath": "params_shard_237.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.53.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "9833da0879b31d506b78bed575b7d634" + }, + { + "dataPath": "params_shard_238.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.53.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "7a7fb367e947a446f395bf65a4264b41" + }, + { + "dataPath": "params_shard_239.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.54.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "164581bfa7adb0698e17055f640749c1" + }, + { + "dataPath": "params_shard_240.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.54.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "c76d03ee42e01c26dc193a636efc52d6" + }, + { + "dataPath": "params_shard_241.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.54.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "6697814b8218f45ce787b5ccfd8271d8" + }, + { + "dataPath": "params_shard_242.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.54.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "57a42b7eb6ba943fc06e3e3230d690d9" + }, + { + "dataPath": "params_shard_243.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.55.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a0d4f276466046b8f7b60170de7ecab6" + }, + { + "dataPath": "params_shard_244.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.55.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "d77a36011e216544ed05d20b282fb29a" + }, + { + "dataPath": "params_shard_245.bin", + "format": "raw-shard", + "nbytes": 26820608, + "records": [ + { + "name": "model.layers.53.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.53.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.53.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.53.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.53.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.54.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9732096 + }, + { + "name": "model.layers.54.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9748480 + }, + { + "name": "model.layers.54.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13418496 + }, + { + "name": "model.layers.54.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.54.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20774912 + }, + { + "name": "model.layers.54.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22085632 + }, + { + "name": "model.layers.55.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23134208 + }, + { + "name": "model.layers.55.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23150592 + } + ], + "md5sum": "c21af6835933bc3cfa3e60789c9e92cc" + }, + { + "dataPath": "params_shard_246.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.55.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "ac1571b98b8a723e8fed09da38a80d03" + }, + { + "dataPath": "params_shard_247.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.55.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "23a362deced8625b3e7ce0a35336b1ed" + }, + { + "dataPath": "params_shard_248.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.56.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "d5f3e94357507c66d8ba67eee4f069fc" + }, + { + "dataPath": "params_shard_249.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.56.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "9a245385ed00b3edefe026b58ee46f5b" + }, + { + "dataPath": "params_shard_250.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.56.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "34e405cf366116d3d3232f3eb6b32d27" + }, + { + "dataPath": "params_shard_251.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.56.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "64d460ce1bc67db09bf9dc334abf2fc4" + }, + { + "dataPath": "params_shard_252.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.57.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "9e1ab277a348d96cf551ffc6532697bf" + }, + { + "dataPath": "params_shard_253.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.57.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "2290c085d7661cb0fc602c7f26ce2805" + }, + { + "dataPath": "params_shard_254.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.55.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.55.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.55.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.55.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.56.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 9715712 + }, + { + "name": "model.layers.56.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 17055744 + }, + { + "name": "model.layers.56.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 18366464 + }, + { + "name": "model.layers.56.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19415040 + }, + { + "name": "model.layers.56.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 19431424 + }, + { + "name": "model.layers.56.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.57.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.57.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "1dd573e366d4d1aaa90080dd7511cca8" + }, + { + "dataPath": "params_shard_255.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.57.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "5fd6e5e03a59dc6882aa33fc925118f5" + }, + { + "dataPath": "params_shard_256.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.57.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "ac143e90513fa52ec1a40e6db638fd6e" + }, + { + "dataPath": "params_shard_257.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.58.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "3f040b4afe56386350cb476ebec5a41a" + }, + { + "dataPath": "params_shard_258.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.58.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a0a49022e96cb7290b496bd06f1a5e0e" + }, + { + "dataPath": "params_shard_259.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.58.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "7c7fe77c7efc850d33529e76a0156cf3" + }, + { + "dataPath": "params_shard_260.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.58.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "cbe6ee8d0418ed0b5b5d16b69accf33f" + }, + { + "dataPath": "params_shard_261.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.59.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "e0aba97917ba48dde339bb06426af42e" + }, + { + "dataPath": "params_shard_262.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.59.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "bf2cfeffeb77a095d63009601838cf34" + }, + { + "dataPath": "params_shard_263.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.59.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "6d43a7d3439da6cafb1d361bbe6134bb" + }, + { + "dataPath": "params_shard_264.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.59.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "b27ad5c4ebb45a785f269a59c36ba4c2" + }, + { + "dataPath": "params_shard_265.bin", + "format": "raw-shard", + "nbytes": 29163520, + "records": [ + { + "name": "model.layers.57.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.57.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.57.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.57.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.58.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.58.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.58.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.58.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.58.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.58.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.59.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.59.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 24428544 + }, + { + "name": "model.layers.59.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.59.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 25493504 + } + ], + "md5sum": "040a74e806927771274aec4c0e2cf88f" + }, + { + "dataPath": "params_shard_266.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.60.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "d2ce586c90737fbf83387e4ff76fd92a" + }, + { + "dataPath": "params_shard_267.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.60.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a40bd4d6236a318ac29496f1dc821e31" + }, + { + "dataPath": "params_shard_268.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.60.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "f469de416d95adc59a3281719d969a35" + }, + { + "dataPath": "params_shard_269.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.60.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "42dce6a4452969649f5782e0b9ff3166" + }, + { + "dataPath": "params_shard_270.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.61.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "9e2506e0735f1619a25391f8cb00a3e3" + }, + { + "dataPath": "params_shard_271.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.61.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "8bf3b76e494ccb04e683ca1dabc8e443" + }, + { + "dataPath": "params_shard_272.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.61.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "23c1554d40d2b1ed09e0619d0396f25e" + }, + { + "dataPath": "params_shard_273.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.61.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "0754a408837c29c712f64941f6d296b4" + }, + { + "dataPath": "params_shard_274.bin", + "format": "raw-shard", + "nbytes": 33112064, + "records": [ + { + "name": "model.layers.59.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.59.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.60.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7356416 + }, + { + "name": "model.layers.60.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 7372800 + }, + { + "name": "model.layers.60.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 11042816 + }, + { + "name": "model.layers.60.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 18382848 + }, + { + "name": "model.layers.60.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 18399232 + }, + { + "name": "model.layers.60.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 19709952 + }, + { + "name": "model.layers.61.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.61.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 20774912 + }, + { + "name": "model.layers.61.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 24444928 + }, + { + "name": "model.layers.61.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31784960 + }, + { + "name": "model.layers.61.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 31801344 + } + ], + "md5sum": "6d322740dfb06699aa2897b6495e3e83" + }, + { + "dataPath": "params_shard_275.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.62.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "12f093124dd78437172959efcc6bab87" + }, + { + "dataPath": "params_shard_276.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.62.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "f45a9e78077a25180edbee909763b257" + }, + { + "dataPath": "params_shard_277.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.62.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a8ca7bfd62c76ce8624bb0d15516c17b" + }, + { + "dataPath": "params_shard_278.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.62.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "475e9eebfddcfdf4a64be602aa5196e4" + }, + { + "dataPath": "params_shard_279.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.63.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a879718f4838d2b26daec19c6aaf7f65" + }, + { + "dataPath": "params_shard_280.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.63.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "320311774811e82e149ec3fbdffaae5d" + }, + { + "dataPath": "params_shard_281.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.63.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "9cec2f41b2e129b8ddbdee11ef9b287c" + }, + { + "dataPath": "params_shard_282.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.63.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "04c24eeab5c6547d0f45f5b452227671" + }, + { + "dataPath": "params_shard_283.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.64.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "a01a98bf8a19f3aec3ea1e6645b36443" + }, + { + "dataPath": "params_shard_284.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.64.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "de47b7cb3f8783f666924d2b647591b7" + }, + { + "dataPath": "params_shard_285.bin", + "format": "raw-shard", + "nbytes": 31522816, + "records": [ + { + "name": "model.layers.61.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 0 + }, + { + "name": "model.layers.62.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 1048576 + }, + { + "name": "model.layers.62.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.62.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2375680 + }, + { + "name": "model.layers.62.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6045696 + }, + { + "name": "model.layers.62.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.62.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 13402112 + }, + { + "name": "model.layers.63.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 14450688 + }, + { + "name": "model.layers.63.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 14467072 + }, + { + "name": "model.layers.63.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 18137088 + }, + { + "name": "model.layers.63.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.63.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 25493504 + }, + { + "name": "model.layers.63.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 26804224 + }, + { + "name": "model.layers.64.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27852800 + } + ], + "md5sum": "6a0fc587e7aa9a936d9b2d345074a412" + }, + { + "dataPath": "params_shard_286.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.64.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "2c1a49c730c166688949063004f4e7ba" + }, + { + "dataPath": "params_shard_287.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.64.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "9a4403ab2b08efb80d0873e0b595f72f" + }, + { + "dataPath": "params_shard_288.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.65.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "7804e4d1930eba5c4ebc93d98e609fbe" + }, + { + "dataPath": "params_shard_289.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.65.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "bb920f4103c85636c28a64eaf952740a" + }, + { + "dataPath": "params_shard_290.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.65.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "0208072e88b70dfbab57dc0b29da0447" + }, + { + "dataPath": "params_shard_291.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.65.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "510cb20c74420eb8e86f7b8930f2a6e3" + }, + { + "dataPath": "params_shard_292.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.66.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "3751901cc60430689dfc19b8b2ada9d2" + }, + { + "dataPath": "params_shard_293.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.66.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "6154a3c7f329880278b3de79a0fb7918" + }, + { + "dataPath": "params_shard_294.bin", + "format": "raw-shard", + "nbytes": 26820608, + "records": [ + { + "name": "model.layers.64.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.64.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.64.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.64.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.64.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.65.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9732096 + }, + { + "name": "model.layers.65.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9748480 + }, + { + "name": "model.layers.65.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13418496 + }, + { + "name": "model.layers.65.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.65.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20774912 + }, + { + "name": "model.layers.65.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22085632 + }, + { + "name": "model.layers.66.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23134208 + }, + { + "name": "model.layers.66.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23150592 + } + ], + "md5sum": "2911ac1d130de49bcebc235c38dc8d9b" + }, + { + "dataPath": "params_shard_295.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.66.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "a19dd0a169114c254a8a11136a36da6f" + }, + { + "dataPath": "params_shard_296.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.66.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "60590b74c97aa21e1a3debdcf6e1218b" + }, + { + "dataPath": "params_shard_297.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.67.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "d0e6e06e256560d0587dde2d0f4b0afc" + }, + { + "dataPath": "params_shard_298.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.67.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "6de03db5143f1f63e009522bbadf22b0" + }, + { + "dataPath": "params_shard_299.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.67.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "687fb46a043e8aba42e4f06ea0773129" + }, + { + "dataPath": "params_shard_300.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.67.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "7eaf55d2bde925dfc1ecbfd89bd93612" + }, + { + "dataPath": "params_shard_301.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.68.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "47f7f8dc18a71628d642c1e5bfa5b53a" + }, + { + "dataPath": "params_shard_302.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.68.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "54d585df688185dc545026632c2e63fc" + }, + { + "dataPath": "params_shard_303.bin", + "format": "raw-shard", + "nbytes": 26804224, + "records": [ + { + "name": "model.layers.66.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.66.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.66.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.66.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.67.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 9715712 + }, + { + "name": "model.layers.67.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 17055744 + }, + { + "name": "model.layers.67.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 18366464 + }, + { + "name": "model.layers.67.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19415040 + }, + { + "name": "model.layers.67.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 19431424 + }, + { + "name": "model.layers.67.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23101440 + }, + { + "name": "model.layers.68.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23117824 + }, + { + "name": "model.layers.68.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23134208 + } + ], + "md5sum": "6bfb50854554c892eada8c1b7f777ce8" + }, + { + "dataPath": "params_shard_304.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.68.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "a58059b379c9005ae7322fc454530436" + }, + { + "dataPath": "params_shard_305.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.68.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "fbaf4d07128fc051dbf86db990cc6437" + }, + { + "dataPath": "params_shard_306.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.69.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "01170ba2cf229d50e0f9fc2aa90a5408" + }, + { + "dataPath": "params_shard_307.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.69.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "198f38682eccf827d22e93ff154c3d63" + }, + { + "dataPath": "params_shard_308.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.69.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "a9799a337c9f35ece22ad559d23890c5" + }, + { + "dataPath": "params_shard_309.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.69.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "63a331fb70d72dadfbb3997d5f1b0477" + }, + { + "dataPath": "params_shard_310.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.70.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "f82792a2b3e2410caf10cc6cf6b764b2" + }, + { + "dataPath": "params_shard_311.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.70.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "01687d6bef0cba14eafbc7dd6f597e0e" + }, + { + "dataPath": "params_shard_312.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "7df0d88a09f348cd89db19ec059a5a00" + }, + { + "dataPath": "params_shard_313.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "4d19f2cef20bf892e8807f8c43218bf9" + }, + { + "dataPath": "params_shard_314.bin", + "format": "raw-shard", + "nbytes": 29163520, + "records": [ + { + "name": "model.layers.68.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.68.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.68.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.68.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.69.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.69.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9732096 + }, + { + "name": "model.layers.69.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13402112 + }, + { + "name": "model.layers.69.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20742144 + }, + { + "name": "model.layers.69.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20758528 + }, + { + "name": "model.layers.69.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22069248 + }, + { + "name": "model.layers.70.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 23117824 + }, + { + "name": "model.layers.70.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 24428544 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 25493504 + } + ], + "md5sum": "7f74dc9d4797abae08abe1b68ed8bc22" + }, + { + "dataPath": "params_shard_315.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c35ec83a75870ac1ad973fc7ea4c5b91" + }, + { + "dataPath": "params_shard_316.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "fc715a7a0e052f8b3c117e44dd5a6e6f" + }, + { + "dataPath": "params_shard_317.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "2c64ac37a1be6ec6d40911a2c5e55934" + }, + { + "dataPath": "params_shard_318.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.8.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "795b5899e57f7f8dab6cf56e9ac04292" + }, + { + "dataPath": "params_shard_319.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "1359ab569a44555eccefb97d9aae48fb" + }, + { + "dataPath": "params_shard_320.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "e05845bccdad9acfebd8dad0f523b212" + }, + { + "dataPath": "params_shard_321.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.9.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "ae5e77d97fad898a10ab76fc09d1db8b" + }, + { + "dataPath": "params_shard_322.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "df2ea91c4a0896b5f1e8b8aa4e1a3645" + }, + { + "dataPath": "params_shard_323.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.70.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "bd0ca84f6bfa6c46115a0f73985947d2" + }, + { + "dataPath": "params_shard_324.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.70.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a3e131ce3fe4c61c13ee4186d12c8482" + }, + { + "dataPath": "params_shard_325.bin", + "format": "raw-shard", + "nbytes": 31522816, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 7356416 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 8404992 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 8421376 + }, + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 12091392 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19431424 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 19447808 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 20758528 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 21807104 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 25477120 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 26787840 + }, + { + "name": "model.layers.70.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 27836416 + }, + { + "name": "model.layers.70.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27852800 + } + ], + "md5sum": "a7a57817bfd8c47a67070c38840c2037" + }, + { + "dataPath": "params_shard_326.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.71.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "d9fcb57996e67ee29e5f333c41d39fe9" + }, + { + "dataPath": "params_shard_327.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.71.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "f56d8f11038586faa5fc026416ab2626" + }, + { + "dataPath": "params_shard_328.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.71.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "175422f002d4557c83742ae442b5420b" + }, + { + "dataPath": "params_shard_329.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.71.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "3af4d0b7868501f4ea5d38c4604148fc" + }, + { + "dataPath": "params_shard_330.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.72.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "5ae1f24811e9c010dc08a730b28811ad" + }, + { + "dataPath": "params_shard_331.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.72.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "b8de7d5fd13807c47cd8399958091ec8" + }, + { + "dataPath": "params_shard_332.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.72.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "cb6f71e0ee6bc62aba118ead3d7f2339" + }, + { + "dataPath": "params_shard_333.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.72.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "0dde7eaee5dd445274d3df6aeac14c7f" + }, + { + "dataPath": "params_shard_334.bin", + "format": "raw-shard", + "nbytes": 33112064, + "records": [ + { + "name": "model.layers.70.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.70.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.71.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7356416 + }, + { + "name": "model.layers.71.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 7372800 + }, + { + "name": "model.layers.71.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 11042816 + }, + { + "name": "model.layers.71.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 18382848 + }, + { + "name": "model.layers.71.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 18399232 + }, + { + "name": "model.layers.71.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 19709952 + }, + { + "name": "model.layers.72.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.72.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 20774912 + }, + { + "name": "model.layers.72.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 24444928 + }, + { + "name": "model.layers.72.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31784960 + }, + { + "name": "model.layers.72.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 31801344 + } + ], + "md5sum": "600362f884b7238be52ccaf31821e46b" + }, + { + "dataPath": "params_shard_335.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.73.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "2bbb8b02db10006e9663ac28952e3c45" + }, + { + "dataPath": "params_shard_336.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.73.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "271975ff3d836e4901ed60d9f718b3a5" + }, + { + "dataPath": "params_shard_337.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.73.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "117dba1221338648979e03a339acccfa" + }, + { + "dataPath": "params_shard_338.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.73.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "46ba3a3f24c50b1783cb237a0b63d061" + }, + { + "dataPath": "params_shard_339.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.74.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "3e7b6bc4e5ecb233ebe4692433434bc0" + }, + { + "dataPath": "params_shard_340.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.74.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "449840957f4f0b551d031a2a597bc3b0" + }, + { + "dataPath": "params_shard_341.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.74.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "7045501cbae7b550760e8c61f95e25af" + }, + { + "dataPath": "params_shard_342.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.74.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "60a5d1fc00b5aa262d54099e6292122f" + }, + { + "dataPath": "params_shard_343.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.75.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "ef66d1981282b5113eee9f78e98dae18" + }, + { + "dataPath": "params_shard_344.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.75.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a7b2741329c44dda25e8b41842bb7755" + }, + { + "dataPath": "params_shard_345.bin", + "format": "raw-shard", + "nbytes": 31522816, + "records": [ + { + "name": "model.layers.72.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 0 + }, + { + "name": "model.layers.73.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 1048576 + }, + { + "name": "model.layers.73.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 2359296 + }, + { + "name": "model.layers.73.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 2375680 + }, + { + "name": "model.layers.73.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 6045696 + }, + { + "name": "model.layers.73.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13385728 + }, + { + "name": "model.layers.73.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 13402112 + }, + { + "name": "model.layers.74.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 14450688 + }, + { + "name": "model.layers.74.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 14467072 + }, + { + "name": "model.layers.74.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 18137088 + }, + { + "name": "model.layers.74.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 25477120 + }, + { + "name": "model.layers.74.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 25493504 + }, + { + "name": "model.layers.74.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 26804224 + }, + { + "name": "model.layers.75.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 27852800 + } + ], + "md5sum": "7741aa5a3a8e68f5e239914ccfec264d" + }, + { + "dataPath": "params_shard_346.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.75.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "b32627d98d7e0b47ecafbc3d7908517c" + }, + { + "dataPath": "params_shard_347.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.75.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "c2ec81cffd0d0c7d611b8ba1c7d7e3bf" + }, + { + "dataPath": "params_shard_348.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.76.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "74580d2fd63d7d54b9ac3d15919c1fab" + }, + { + "dataPath": "params_shard_349.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.76.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "3d6c7cfc8a9f336426b551ab678412a3" + }, + { + "dataPath": "params_shard_350.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.76.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "0197888c16cff0cd3a6a5586067e3420" + }, + { + "dataPath": "params_shard_351.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.76.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "4143d1ad630498e4e55e35d2cbd13137" + }, + { + "dataPath": "params_shard_352.bin", + "format": "raw-shard", + "nbytes": 117440512, + "records": [ + { + "name": "model.layers.77.mlp.down_proj.q_weight", + "shape": [ + 8192, + 3584 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 117440512, + "byteOffset": 0 + } + ], + "md5sum": "0938442b6a60f300a5e293172940586d" + }, + { + "dataPath": "params_shard_353.bin", + "format": "raw-shard", + "nbytes": 234881024, + "records": [ + { + "name": "model.layers.77.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 234881024, + "byteOffset": 0 + } + ], + "md5sum": "a2c4226e9c5b4c11726f3c55a75c32e0" + }, + { + "dataPath": "params_shard_354.bin", + "format": "raw-shard", + "nbytes": 26820608, + "records": [ + { + "name": "model.layers.75.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.75.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7340032 + }, + { + "name": "model.layers.75.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8650752 + }, + { + "name": "model.layers.75.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9699328 + }, + { + "name": "model.layers.75.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9715712 + }, + { + "name": "model.layers.76.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 9732096 + }, + { + "name": "model.layers.76.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 9748480 + }, + { + "name": "model.layers.76.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 13418496 + }, + { + "name": "model.layers.76.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 20758528 + }, + { + "name": "model.layers.76.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 20774912 + }, + { + "name": "model.layers.76.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 22085632 + }, + { + "name": "model.layers.77.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23134208 + }, + { + "name": "model.layers.77.mlp.down_proj.q_scale", + "shape": [ + 8192, + 224 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3670016, + "byteOffset": 23150592 + } + ], + "md5sum": "7bc67155ff807300671f3e85bc524e3f" + }, + { + "dataPath": "params_shard_355.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.77.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "ca8d0b38868fc8875d04c270f5d9cba7" + }, + { + "dataPath": "params_shard_356.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.77.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "09957de05cc80a3691b32ce962886cc9" + }, + { + "dataPath": "params_shard_357.bin", + "format": "raw-shard", + "nbytes": 41943040, + "records": [ + { + "name": "model.layers.78.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 41943040, + "byteOffset": 0 + } + ], + "md5sum": "867b325d751542a567003c5d0b217bb3" + }, + { + "dataPath": "params_shard_358.bin", + "format": "raw-shard", + "nbytes": 33554432, + "records": [ + { + "name": "model.layers.78.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 1024 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33554432, + "byteOffset": 0 + } + ], + "md5sum": "5a4ed8b5328ffaa91a0276bb67dc7de4" + }, + { + "dataPath": "params_shard_359.bin", + "format": "raw-shard", + "nbytes": 12075008, + "records": [ + { + "name": "model.layers.77.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 0 + }, + { + "name": "model.layers.77.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 7340032 + }, + { + "name": "model.layers.77.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 7356416 + }, + { + "name": "model.layers.77.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 8667136 + }, + { + "name": "model.layers.78.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1310720, + "byteOffset": 9715712 + }, + { + "name": "model.layers.78.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1048576, + "byteOffset": 11026432 + } + ], + "md5sum": "2e2a30b0f07cd79a764420fb2e26fd6a" + } + ] +} \ No newline at end of file