diff --git "a/ndarray-cache.json" "b/ndarray-cache.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache.json" @@ -0,0 +1,12343 @@ +{ + "metadata": { + "ParamSize": 805, + "ParamBytes": 31776318464.0, + "BitsPerParam": 3.04023285660184 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 420679680, + "records": [ + { + "name": "lm_head.q_weight", + "shape": [ + 128256, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 420679680, + "byteOffset": 0 + } + ], + "md5sum": "9cf16ebcdd3d9a19e4049558f6b42e90" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 52584960, + "records": [ + { + "name": "lm_head.q_scale", + "shape": [ + 128256, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 52584960, + "byteOffset": 0 + } + ], + "md5sum": "1f30e44d9b43b26ca88956cdbfa05437" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.78.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "c8f48cea1c953e182c8f0ffef06d4238" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.79.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "481d39272d969a99061589a30d9fe5a7" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 23560192, + "records": [ + { + "name": "model.layers.78.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 0 + }, + { + "name": "model.layers.78.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 16384 + }, + { + "name": "model.layers.78.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.79.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23543808 + } + ], + "md5sum": "8d9aa2da7ef0d186919694290fa169d3" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.79.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "fd8c1ca4e556368cd1fa4ce83471eafb" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.79.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "6dc699d36850fe5d9bf7d79880e6a7e7" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.79.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "fc6a4bce2235d8767ddbe612003a4180" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.79.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a1282147152a068b385f25b4f40bf103" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 420679680, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 128256, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 420679680, + "byteOffset": 0 + } + ], + "md5sum": "4b63bbd661056dbef8cb4b022f7e9d80" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 52584960, + "records": [ + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 128256, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 52584960, + "byteOffset": 0 + } + ], + "md5sum": "0e3738fae1222d3b78775b5b399caacb" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "88c54d298d2ab932d67bf57f919a3602" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "0c34ee932cba906439f0e3e7bc3effa1" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "3bbe69ffb6c64012c12958f73820d779" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.0.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "23ad995e3f0a1770bc31077cd46b98be" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.79.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.79.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.79.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.79.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.norm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "6163be61d9e07c65f01c99c4e5ad4539" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "12fcd2a3eaf02baeedb07b5a4ab25769" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "3c1703af70e72a003f81e2a806c47e68" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.1.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "e219f8f9edf4adcafea76b94d010ad1d" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "5a84bebe6da3fefac7ebcfe8db9e3d2a" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "9632665694a636664832e0c7bb700a44" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "2b200e951b3815008fa98cfeae9a540d" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "fbba44257a548692a5270108d530f974" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "5d53eed196aefc4a231f346d6afe91d1" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "bf5a46f5e327561fb39d36ea09dcea85" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.2.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "6170fea86768bcc39743eae2847e8fba" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "983c6d17409074bdfb11dabefb99dcf6" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15138816 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15155200 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26902528 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26918912 + } + ], + "md5sum": "cb494fee1dad306dbe952dc37a78257f" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "f36910f6711d3d0f2ebbfc7595fc216e" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "dae2f8881fede83c20471b9f3ab683b6" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "f93d35ac94de62abb6bd7e29c2cfed0b" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.3.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "48e6ef5def924c93bbe07bb3c2b7d4c9" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "ef0f0d783bf9d0abff69b9265298b9b9" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.4.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "7f36d4b540b023c464a6a3b92f013430" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "e901f4e255af0cc734fc8c7946fa8057" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "4274fd6a80f71e57c7c3b6b617e9a917" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "dc490fe2a1c7ad67a31148350e71968a" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "c1d882ac5a3c4cb7dc08d19fb0195350" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "019456d7666b2a63e43b4fde92715264" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.10.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "53d140d73e1aa15b047ffc42e99e60cd" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "0f2f6a53d1bbc4e0fddd8ba9b4e62532" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "5ddfa7cd855ec86b7b5dd4d458eec3a7" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "67328f7a926cfad5336e4d1651414a8b" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "1a77f2183e8b740cb47ea08e5698c7b1" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.11.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5aa0a4603c926f68b485835bd8608023" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "0fe7b43c1983e3ccf2c3db03b9857559" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "15fa27c3da7bc629fbc370b7a0a2affa" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.12.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "a9d83ed4c83d2ae2571d60767fd63e8a" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "cb6a60c6ecebccaef4dcc815f690795a" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "95f74584ad503266fa9f7c0a06071bd7" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f857e5c612e0140b3bc67f0abbe00231" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "689515643766b61043e13b2666d6a561" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "c2848987ddd835db9b484f89a4afd9cf" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 3358720 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 7557120 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 10915840 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 10932224 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22679552 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "9c9ddbbab82f5954113e6ceb84460770" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "3dee722251b69ac4dff7f5de78b49824" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "30c981645305bacedd42c39fc08a3fb5" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "84733bf110ed2264357c7c4fe772e0a1" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "a02157cdc77a90869f482fcf2afe14f1" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "9d522f94f2caef87da90b112b034d0fd" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.13.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0c50b0cf4126747af7627c2ad5771b5d" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "efbf4781132182d8dbea8e345cb5257d" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "52837043edbb1f78adfc94444e737631" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "203ed28df11deb72978d9402147be509" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "b9fa8e69282c7442efd91aba597e260d" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "9424ca1dde3c36b4221f8ed1c2c55388" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.14.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ec8fa57e8079771c08f55c85e9e9647a" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "92b386b4fce190f2effc47377c8315cb" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "d381d3f4c974e949c07bbc07dd27c42e" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ff110a6c15460ec16472d44c8b873629" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "fc9f5f681a8f2d5c27dbe3016cfeb28c" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.15.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "14836640bb910ab7cca8c3210fcfafa0" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "61d9f3070bae76a54814e05773408795" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "0ddc2a84f62d2c41cd116a31a110e5cd" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "f0aa4172d5451e98679c736b1e3c42c6" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f7e9f82f7b194e038482b3f409b21add" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "68e2c3a15e8bb88a32b56170037a6201" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.16.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "1d15b082c7949435997d2d34e59ed10a" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "c00f3a4cde4f2b70866efd049d24e2d7" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ba302ed4be03f2894304ec1b2815e80b" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "2fcae42f4aa2b7811ffaac0f892ab1ca" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.17.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "a9d5bb17840ae4e63664708069bf0113" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "b64554ae5dd3de3c245cda978b7be361" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "fe2a4cf57630aae1dfaaf36d4993359a" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "9b3f1ad916072a99aeeeb9f1840703bf" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "9bdcd0bce15d58d2bd9f496d9da08c79" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "6c2321fd5231794af3dfbefc82cf0568" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "82163c24f6a13b390306b1b14b2fe7aa" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.18.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5d38ab0e6055b2293c84856aaf79a93f" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "b38918ee9f75a7bf06fd5adb10e788ce" + }, + { + "dataPath": "params_shard_89.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "fdb5e31dbd7baa85d6a8899cb2cade50" + }, + { + "dataPath": "params_shard_90.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "9546ba429cb9285f8ef563f77494291b" + }, + { + "dataPath": "params_shard_91.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ae4dbae0d646fe02e1afd1bab8e26150" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "09fb17963360bfc4f6c3175c4e5a86df" + }, + { + "dataPath": "params_shard_93.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.19.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "185843ac28fe4afb3091d9f80ca650ae" + }, + { + "dataPath": "params_shard_94.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "7d208455888415f74eeca206d47d8d4a" + }, + { + "dataPath": "params_shard_95.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.20.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ea4391fe2e1a871f194be2f9b8e59ceb" + }, + { + "dataPath": "params_shard_96.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "e36757b76c2338b50c071a380e70f0f3" + }, + { + "dataPath": "params_shard_97.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "a0df39acefaae67624ea23089552b324" + }, + { + "dataPath": "params_shard_98.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "2b3551247fdab08de6ee16cea5f63e60" + }, + { + "dataPath": "params_shard_99.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "5ea773e0007078030ca061891a1ada5c" + }, + { + "dataPath": "params_shard_100.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "418d40b0ca371fe87cdbfb260fe01637" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "7dea13c5eb155af994d95e29fc2907c9" + }, + { + "dataPath": "params_shard_102.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "a1570eadf3d98a5eeee5d97e3708e90c" + }, + { + "dataPath": "params_shard_103.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "cec37ff63a556f1903d345785cd585d6" + }, + { + "dataPath": "params_shard_104.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.21.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "b3e6054f1f3caa8aa8528a40e7dafd2e" + }, + { + "dataPath": "params_shard_105.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "125663124f0e333b70f220e04ff1063b" + }, + { + "dataPath": "params_shard_106.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "c718853f2512532e75536b83dbee71c3" + }, + { + "dataPath": "params_shard_107.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "1c2affbad59156ce5687e6fbd197a285" + }, + { + "dataPath": "params_shard_108.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "5d311c12e4ca725cc223cf40888fac95" + }, + { + "dataPath": "params_shard_109.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "3807c6ea3bc5676bc27ea0fa8fd1500a" + }, + { + "dataPath": "params_shard_110.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.22.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5b30bfcb66056a8f7b004cb62c89d770" + }, + { + "dataPath": "params_shard_111.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "3589b56fb0ba790408ebcda25acb1050" + }, + { + "dataPath": "params_shard_112.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "0a273720a6a5bc812020c9e44a23b4c5" + }, + { + "dataPath": "params_shard_113.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "1d67e2d69c542b73f082f1624a3244cb" + }, + { + "dataPath": "params_shard_114.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "82a304d0b8d88aac812a8f2324a04f78" + }, + { + "dataPath": "params_shard_115.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.23.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "27049565b6134bff574c381bf61d2829" + }, + { + "dataPath": "params_shard_116.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "06bc26b29bcfdea8f9a3854520b414f6" + }, + { + "dataPath": "params_shard_117.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "26779675de91a46860e8a444352c172f" + }, + { + "dataPath": "params_shard_118.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "d841e318e056c3ed943e167b99029edd" + }, + { + "dataPath": "params_shard_119.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ec791a8d97b7a3078f40cf213df72659" + }, + { + "dataPath": "params_shard_120.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "5f7ad7347f66c6fb2c02ef97f2d67c92" + }, + { + "dataPath": "params_shard_121.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.24.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "454b654f87206579cf2ee11279c851d1" + }, + { + "dataPath": "params_shard_122.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a0c3844bdcb5fef309d035ab6ae4afeb" + }, + { + "dataPath": "params_shard_123.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "e4a0aaf57d1af0eb2440fee4475f682b" + }, + { + "dataPath": "params_shard_124.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "a451712ded80478d59e35eb819414906" + }, + { + "dataPath": "params_shard_125.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.25.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "b84e18222abc988882876206b42ccabd" + }, + { + "dataPath": "params_shard_126.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "d9bcec59a823b40eba5d23862f076cb3" + }, + { + "dataPath": "params_shard_127.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "ff8f0ab682ddd4a6035af482ffa6e924" + }, + { + "dataPath": "params_shard_128.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "4df3051a5a07c3768be62d7df31180d1" + }, + { + "dataPath": "params_shard_129.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "43dc27f76d4e00816415bfd99a93f8ff" + }, + { + "dataPath": "params_shard_130.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ef6d8185b08409d70bdcaa867f6fbc23" + }, + { + "dataPath": "params_shard_131.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "33cbc143e18aaed8abfc7e3aece4d5f5" + }, + { + "dataPath": "params_shard_132.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.26.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "55c3458730b8f023186855ab1cdec443" + }, + { + "dataPath": "params_shard_133.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "36ea4ac92462dc4779b31d67276af488" + }, + { + "dataPath": "params_shard_134.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "7405e9eca0cc1276b49409f87012aee6" + }, + { + "dataPath": "params_shard_135.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "f0e202da4bde4fc33fa4677ecd4ff932" + }, + { + "dataPath": "params_shard_136.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "652ce4a84398a4e0c0f8890cd36121f8" + }, + { + "dataPath": "params_shard_137.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "7ea6108f35e6d1a3a00a6bf874e84fd2" + }, + { + "dataPath": "params_shard_138.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.27.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "025a616136bcc044cf20d68a993d00d7" + }, + { + "dataPath": "params_shard_139.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "0532bc8f6e58fa6e33803a31f6475de8" + }, + { + "dataPath": "params_shard_140.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.28.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "05ad08cb79931858799db7c0d8417ce2" + }, + { + "dataPath": "params_shard_141.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "4692cdf47131b60f157ae16fd483e405" + }, + { + "dataPath": "params_shard_142.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "54503d0a10a792af02aeb1e2ce4fc56a" + }, + { + "dataPath": "params_shard_143.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "04ac4037dc551b0327594d765db6651a" + }, + { + "dataPath": "params_shard_144.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "a96597d2963718bd76c64ef87217c7ed" + }, + { + "dataPath": "params_shard_145.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "4a45df3cd00fd584a24c8be9c4f35c14" + }, + { + "dataPath": "params_shard_146.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "cda33e0860c00da6b040bd1c566c0669" + }, + { + "dataPath": "params_shard_147.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "9e85e3451738f8c49ec6032a66ae6289" + }, + { + "dataPath": "params_shard_148.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "291ee41e446dbb0863ef0bed298801d9" + }, + { + "dataPath": "params_shard_149.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.29.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "922270f50cef1e83b5cb8b3afcc758a0" + }, + { + "dataPath": "params_shard_150.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "420629fe011af2fd32ba6ab9c5c3661e" + }, + { + "dataPath": "params_shard_151.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "121bcd1e31f186a71555fa795493b7f8" + }, + { + "dataPath": "params_shard_152.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "a08e042a4fc8f12f3fa2822ce98a6855" + }, + { + "dataPath": "params_shard_153.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f1d295d70edf4d1f15e8e81f21d08f60" + }, + { + "dataPath": "params_shard_154.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "83e8a8881e94f08b8d92f5269d8b0e5a" + }, + { + "dataPath": "params_shard_155.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.30.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "9f8ab9942a7eba8eeef7ba1121e72f81" + }, + { + "dataPath": "params_shard_156.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "c5ed65168ab175f1abec8e77a8b9ecfb" + }, + { + "dataPath": "params_shard_157.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "9bfca638cc6c6989da30923b28babd37" + }, + { + "dataPath": "params_shard_158.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "757f74c445e592052596328a9ba19180" + }, + { + "dataPath": "params_shard_159.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "62c5fc1fcc653db981b57b91e7d369a3" + }, + { + "dataPath": "params_shard_160.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.31.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "cd263704e5c6e77aeff89d9e0d4742ef" + }, + { + "dataPath": "params_shard_161.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "45818221462bee619a13ebd9915bcd48" + }, + { + "dataPath": "params_shard_162.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "a2ace1cd44dd0e83794851569f8709d5" + }, + { + "dataPath": "params_shard_163.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.32.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "ab01fad2c270a766a6ce761dffe7a349" + }, + { + "dataPath": "params_shard_164.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "faebfb5b1f7fe209910d7ce560fe6bd3" + }, + { + "dataPath": "params_shard_165.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "c129e37e98dcda4cb51380c0576a4f40" + }, + { + "dataPath": "params_shard_166.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.32.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "df4ff64f6912816b450abe786884adf5" + }, + { + "dataPath": "params_shard_167.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.32.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "ce05f2952adff16a005ffbc8f1adbcf4" + }, + { + "dataPath": "params_shard_168.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "29a662089ec1c9353bb7b72650388d02" + }, + { + "dataPath": "params_shard_169.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "32585fbcf689e2cd039db5fba29bff0d" + }, + { + "dataPath": "params_shard_170.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.33.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "413abb8e200e05659e6b8712c280e0af" + }, + { + "dataPath": "params_shard_171.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.33.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "901c1d6d18b2d4e20031837720257ac7" + }, + { + "dataPath": "params_shard_172.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "ba5345576117d875b08b2a249c731744" + }, + { + "dataPath": "params_shard_173.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "0971cc5d45557bc71bf0bd25cf399951" + }, + { + "dataPath": "params_shard_174.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.34.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "88f4a60522dfaf77cde86868a8c719a7" + }, + { + "dataPath": "params_shard_175.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "772b1e29e49d24dd8b5b48513d398ae6" + }, + { + "dataPath": "params_shard_176.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "1aea52702fbba818ab084b83ac7a6632" + }, + { + "dataPath": "params_shard_177.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.34.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "70dfc729267537eab5957d8ca3e31054" + }, + { + "dataPath": "params_shard_178.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.34.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "21d65248969558ba2d63aec8b71b771c" + }, + { + "dataPath": "params_shard_179.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "5cb5589b3fdb4737876c9b6df0ffb383" + }, + { + "dataPath": "params_shard_180.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.34.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "24ce607df219fdae7250721f9bc77e3b" + }, + { + "dataPath": "params_shard_181.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.35.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "9b266efbccafe7f56bdff44f92cfba01" + }, + { + "dataPath": "params_shard_182.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "31a45ff57603ae8c1dd87766951c1ec9" + }, + { + "dataPath": "params_shard_183.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.35.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "dc6ae8fec930b703ef8bf2d545f2b552" + }, + { + "dataPath": "params_shard_184.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.35.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "1397d6e888e29857c3b3ab454bf1a1fc" + }, + { + "dataPath": "params_shard_185.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.36.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0c654c5a370bbe9abb5e9ce442d22e6b" + }, + { + "dataPath": "params_shard_186.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.36.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "9b120821604b4fa20eb03f84f88852fe" + }, + { + "dataPath": "params_shard_187.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.36.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "61a0b8771d4a50547abe7eceaf75fe59" + }, + { + "dataPath": "params_shard_188.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.36.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "43ddb73ea46996f7c86611763d76f2c3" + }, + { + "dataPath": "params_shard_189.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "65663c013739964a03586d5b454fd9be" + }, + { + "dataPath": "params_shard_190.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "ec4de4eae3acc1b0f4fa5db0432bad8f" + }, + { + "dataPath": "params_shard_191.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "ba0b8b10f81fd62b98e8d67922eb9462" + }, + { + "dataPath": "params_shard_192.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.37.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "16a2f3ee60f241edb2eda709cc91b535" + }, + { + "dataPath": "params_shard_193.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.37.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "8b0b86807f2dc9a03e78429d8f0e071c" + }, + { + "dataPath": "params_shard_194.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.37.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "93d0694f71bbb08b808dab643230aa29" + }, + { + "dataPath": "params_shard_195.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.37.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "b3025f219003466e7e9951bc07cb02d9" + }, + { + "dataPath": "params_shard_196.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.38.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "ed921b9f11aa3930e44b93d07c0b4213" + }, + { + "dataPath": "params_shard_197.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.36.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.36.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.37.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.37.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.37.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.37.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "a3ec9a238ebdc5a640f1a130b35022a7" + }, + { + "dataPath": "params_shard_198.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "a9f3ef8daccf813b0be5715aefeda4d5" + }, + { + "dataPath": "params_shard_199.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "78ea3fd3bacc1a2a64877e46ea66be6b" + }, + { + "dataPath": "params_shard_200.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.38.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "e14f8588b00e2f75128d54616ebbd630" + }, + { + "dataPath": "params_shard_201.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.38.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "0f5e9b39ff9296b9b7b2a4a686a1384f" + }, + { + "dataPath": "params_shard_202.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "18965668e6c23008376ccfd40948af06" + }, + { + "dataPath": "params_shard_203.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.39.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "1e6bf1f2662bfe02d7158ecaad93b4aa" + }, + { + "dataPath": "params_shard_204.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.39.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "a682c5c99c980a3ffd83b16f196ad527" + }, + { + "dataPath": "params_shard_205.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.39.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "f2c0267a611cae3fa83a808b9b4618e4" + }, + { + "dataPath": "params_shard_206.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.38.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.38.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.38.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.39.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.39.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.39.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "aec631dfd5a6a5839854999bc0eb44a3" + }, + { + "dataPath": "params_shard_207.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.39.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.39.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "fcc20c232ec60fca20f3d063b46927c8" + }, + { + "dataPath": "params_shard_208.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "645ab0ad6b1c786cbd5200c391e99362" + }, + { + "dataPath": "params_shard_209.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.40.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "51aa0ddc3ea174ce0bff050e87bf42e0" + }, + { + "dataPath": "params_shard_210.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.40.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "46ed7ba15b020ea56d380c59f840e31f" + }, + { + "dataPath": "params_shard_211.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.40.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "097d7c729ca47d55ac7d9c0bfec515c5" + }, + { + "dataPath": "params_shard_212.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.40.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "23a53bd4763eb76e2bf59371b5ea26d3" + }, + { + "dataPath": "params_shard_213.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "5e837018a93035759c86bdd560320c60" + }, + { + "dataPath": "params_shard_214.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "f72d2a386f8b9d576f495effc73ca38f" + }, + { + "dataPath": "params_shard_215.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.41.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "9d3309fdb14398a9dc4d9a6106895259" + }, + { + "dataPath": "params_shard_216.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.41.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "50a165610fa43e4e2e21f86930086800" + }, + { + "dataPath": "params_shard_217.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "134cce50da37722c81b0d2104711a05f" + }, + { + "dataPath": "params_shard_218.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.39.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.40.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.40.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.40.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.40.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "3753a69687c2c1e4538b25e6d9508b7a" + }, + { + "dataPath": "params_shard_219.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "fe3e1f6c672e4227787eadc117d228ad" + }, + { + "dataPath": "params_shard_220.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "c7751359be7b69cafd5cf7df7f894523" + }, + { + "dataPath": "params_shard_221.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "a1b237478aba6feffb3c4fbc7b59cb0b" + }, + { + "dataPath": "params_shard_222.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "dae4a4740a236d02ef400b25d8560cfd" + }, + { + "dataPath": "params_shard_223.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "f1b4ece36b60b8aa7c7fb2d531354d21" + }, + { + "dataPath": "params_shard_224.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.5.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "4972798a24eba0019abb9a9ab3ad7bdc" + }, + { + "dataPath": "params_shard_225.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "90791285a5cf746c677a161b34e29696" + }, + { + "dataPath": "params_shard_226.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "257e41e7b42266b92d2607fa7ffd94b5" + }, + { + "dataPath": "params_shard_227.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "edf2be170a3bd098913be7636f06bfde" + }, + { + "dataPath": "params_shard_228.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "275bdde986225b3e326644f109c81faa" + }, + { + "dataPath": "params_shard_229.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "4f98ef8809b3d7c9aa78bdf2c5df9792" + }, + { + "dataPath": "params_shard_230.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.6.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "c3f262e54a3aea385f6068a122220593" + }, + { + "dataPath": "params_shard_231.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "28e9d23a03114f08969eefd4e03f3b49" + }, + { + "dataPath": "params_shard_232.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.41.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "13b67034cb8f30b0bda40f3e9317c92d" + }, + { + "dataPath": "params_shard_233.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.42.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "37b50fd19262a876f41a3cca4fc9dd14" + }, + { + "dataPath": "params_shard_234.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.41.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.41.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19320832 + }, + { + "name": "model.layers.41.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31068160 + }, + { + "name": "model.layers.42.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "4d9ebe3c20b3683a34a3aff1a2355f04" + }, + { + "dataPath": "params_shard_235.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.42.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "610ccac534fd1b5dd06a87d0adac832f" + }, + { + "dataPath": "params_shard_236.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.42.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "225ffca39bd380b132e05435fdd8fd25" + }, + { + "dataPath": "params_shard_237.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.42.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ec8e913cf0740fc216be02696fd55c51" + }, + { + "dataPath": "params_shard_238.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.42.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "03ba71c6dd4182a4975a85c36e13eca3" + }, + { + "dataPath": "params_shard_239.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.43.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "abc40efdd62b446d3fbabfa3744cd88f" + }, + { + "dataPath": "params_shard_240.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.43.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "4eb91d62dfd52c207264d4c27c6845a5" + }, + { + "dataPath": "params_shard_241.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.43.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "d2c35b5a1ddd9ac11b8297694f8bb18c" + }, + { + "dataPath": "params_shard_242.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.43.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ba3ebeffc900f9361e54e67c52f42648" + }, + { + "dataPath": "params_shard_243.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.42.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.42.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.42.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.42.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.43.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.43.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.43.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "1bc6a228316f2a2fcd15c314281e624d" + }, + { + "dataPath": "params_shard_244.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.43.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.43.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "afdffe3a88a18543095e06e0dfcfa0f0" + }, + { + "dataPath": "params_shard_245.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.44.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "78d0873363122e267f794629ef31633b" + }, + { + "dataPath": "params_shard_246.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.44.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "54f265b7494c59d22438e721ebe9cf17" + }, + { + "dataPath": "params_shard_247.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.44.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "6b5bbd118b7d0bbfbb7aaf47453f6c61" + }, + { + "dataPath": "params_shard_248.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.44.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "6e8a4eef1e8bc7d6209d5193956cd8ac" + }, + { + "dataPath": "params_shard_249.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.44.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "1077645a179bcfe49cd8b2145c04f98b" + }, + { + "dataPath": "params_shard_250.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.45.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "80f9cd6a5837994daa003ca4525cef43" + }, + { + "dataPath": "params_shard_251.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.43.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.44.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 3358720 + }, + { + "name": "model.layers.44.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 7557120 + }, + { + "name": "model.layers.44.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 10915840 + }, + { + "name": "model.layers.44.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 10932224 + }, + { + "name": "model.layers.44.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22679552 + }, + { + "name": "model.layers.45.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "ee766ff203513f9d47cf515b6ae4d7ff" + }, + { + "dataPath": "params_shard_252.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.45.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "49c1a36148e45e8363fd5121e59e2b8f" + }, + { + "dataPath": "params_shard_253.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.45.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "e087e6ed3a8987bb5d7492c317d75b94" + }, + { + "dataPath": "params_shard_254.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.45.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "78b0819bffdda78f849c2b2a0a935acb" + }, + { + "dataPath": "params_shard_255.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.45.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "cf370141bdb1c510ebf1b2f96c2f9388" + }, + { + "dataPath": "params_shard_256.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.46.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "1853687236262f3505554adc3d7e7ae6" + }, + { + "dataPath": "params_shard_257.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.46.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f458060769f19577824e9bb38e71df1e" + }, + { + "dataPath": "params_shard_258.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.46.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "ea8485b2f9f9caf815fcf5de9b2cc1ce" + }, + { + "dataPath": "params_shard_259.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.46.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "7b29431d9583d087ca1e68ccc8420957" + }, + { + "dataPath": "params_shard_260.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.45.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.45.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.45.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.45.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.46.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19320832 + } + ], + "md5sum": "74eb222e2d9c828d9473d3f4ec21b599" + }, + { + "dataPath": "params_shard_261.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.46.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.46.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "8492199d952bafafec037b80ae61ed40" + }, + { + "dataPath": "params_shard_262.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.47.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "a167d433c10c27e0b9b7a6ef24010652" + }, + { + "dataPath": "params_shard_263.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.47.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "1301723aa96f70bb0b0119dcaae2c156" + }, + { + "dataPath": "params_shard_264.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.47.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "102c6127b5b848c1af95088aa892a6a3" + }, + { + "dataPath": "params_shard_265.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.47.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "24a8ff819cd21e192fa5e7b14a8d7655" + }, + { + "dataPath": "params_shard_266.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.47.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "8984fb9fe49d426d9077cb15d5ac991f" + }, + { + "dataPath": "params_shard_267.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.48.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "683c464767e58c14057d28f48e47ef47" + }, + { + "dataPath": "params_shard_268.bin", + "format": "raw-shard", + "nbytes": 22745088, + "records": [ + { + "name": "model.layers.46.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.46.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.46.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3375104 + }, + { + "name": "model.layers.47.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3391488 + }, + { + "name": "model.layers.47.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3407872 + }, + { + "name": "model.layers.47.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15155200 + }, + { + "name": "model.layers.47.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15171584 + }, + { + "name": "model.layers.47.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19369984 + }, + { + "name": "model.layers.48.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22728704 + } + ], + "md5sum": "bd9d70262ff6be8c42a342ce411aa0c1" + }, + { + "dataPath": "params_shard_269.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.48.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "02177aee68ebe0aa0b3e3c1633a305a9" + }, + { + "dataPath": "params_shard_270.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.48.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "5ed6f8e53883ee7ea70365da7578f569" + }, + { + "dataPath": "params_shard_271.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.48.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0c4292f410952a89f0fa1f4bb8d48941" + }, + { + "dataPath": "params_shard_272.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.48.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "80b1af8f285ae28a3d8fcddd6a882e43" + }, + { + "dataPath": "params_shard_273.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.49.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "3a63c818f74a2cbb6db1291557a97922" + }, + { + "dataPath": "params_shard_274.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.49.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "a4c31911e705c0b47b503d3eef94d20a" + }, + { + "dataPath": "params_shard_275.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.49.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "8f89f42d4749629fdfbc94b4fdb33cd3" + }, + { + "dataPath": "params_shard_276.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.49.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "70ba58f4bd0f4dd6484eef525745cc2d" + }, + { + "dataPath": "params_shard_277.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.49.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "9238caef703e872451b6239f8d6a95b9" + }, + { + "dataPath": "params_shard_278.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.48.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.48.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.48.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.48.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.49.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.49.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.49.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "2124ae22c61719dd26acf8b27e4b304f" + }, + { + "dataPath": "params_shard_279.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.50.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "e12db8b719d1c5771516fcc702b844b1" + }, + { + "dataPath": "params_shard_280.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.50.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "947fd5389158a2ca97eec8b278c2c120" + }, + { + "dataPath": "params_shard_281.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.50.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "25fe46ee97399093a6460652d84a57f2" + }, + { + "dataPath": "params_shard_282.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.50.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "de94279c46928b9df0eddaf440271846" + }, + { + "dataPath": "params_shard_283.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.50.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "d6f390c709f920269bc85b39b95f3bc0" + }, + { + "dataPath": "params_shard_284.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.51.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "b4727dd22cd684dce907cc7b4b026c09" + }, + { + "dataPath": "params_shard_285.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.49.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.49.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.50.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.50.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.50.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.50.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.50.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.51.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "56848387da59990e61f8bfedb18dc614" + }, + { + "dataPath": "params_shard_286.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.51.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "d3cae51e985391520677d2688301a51a" + }, + { + "dataPath": "params_shard_287.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.51.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "467d324252d79ed36471928cd378e14a" + }, + { + "dataPath": "params_shard_288.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.51.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "75e28330e7ea1ddc753440785a0bd12a" + }, + { + "dataPath": "params_shard_289.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.51.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "41b05a9d9242987e82819d2ffa19e9e9" + }, + { + "dataPath": "params_shard_290.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.52.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "bff0160eb0b829e8f3814c73685bb04c" + }, + { + "dataPath": "params_shard_291.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.52.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "88cd69d1597dc741ac467cd6a9fcbc24" + }, + { + "dataPath": "params_shard_292.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.52.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "a17d8f24f01b1aad553c51c1fec00494" + }, + { + "dataPath": "params_shard_293.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.51.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.51.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.51.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.51.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.52.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.52.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.52.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "88f055b61aaf61a403d06cf8d4c2a119" + }, + { + "dataPath": "params_shard_294.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.52.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "a69a058dbd2fccdce9dce543526fdf53" + }, + { + "dataPath": "params_shard_295.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.52.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "854028ea616bdbe5cb7f5dabfe964f07" + }, + { + "dataPath": "params_shard_296.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.53.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "9aabb1898afd0b8c9cd34897355140c3" + }, + { + "dataPath": "params_shard_297.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.53.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "0e7404687b5002d7447e2e508ae40cdb" + }, + { + "dataPath": "params_shard_298.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.53.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "8c185b1431b298cd87345321e1d8c5b1" + }, + { + "dataPath": "params_shard_299.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.53.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "3af4129d6d52d14d171fbd816ccf757c" + }, + { + "dataPath": "params_shard_300.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.53.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "26121ea1edd6db5409533b97a568267b" + }, + { + "dataPath": "params_shard_301.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.54.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "cd74de2dc9df067670359b9f14b3eece" + }, + { + "dataPath": "params_shard_302.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.52.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.52.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.53.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.53.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.53.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.53.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.53.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "49fd10c35f42374b71227eb61de39865" + }, + { + "dataPath": "params_shard_303.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.54.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "b5da04eb808c597fcfbbd64b604f4b2e" + }, + { + "dataPath": "params_shard_304.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.54.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "9f442a1bce863459c2c565056b2ec52b" + }, + { + "dataPath": "params_shard_305.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.54.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0630ab3e135b0430440c0d8476f5a9a8" + }, + { + "dataPath": "params_shard_306.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.54.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a108fa86396cb059856072ec96bc2237" + }, + { + "dataPath": "params_shard_307.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.55.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "f17703d78454dbdd5f59c41c61fc5f80" + }, + { + "dataPath": "params_shard_308.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.55.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "535662148ba7edd777c81638a90c1d81" + }, + { + "dataPath": "params_shard_309.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.55.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "c22b6bc78a2a54295b0bb9fe72c4629a" + }, + { + "dataPath": "params_shard_310.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.55.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "51ac3cd008ada7254c76c5d32cf46780" + }, + { + "dataPath": "params_shard_311.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.54.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.54.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.54.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.54.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.54.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.55.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.55.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.55.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "5e5f495e7da227f0890bcc0261c18ad6" + }, + { + "dataPath": "params_shard_312.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.55.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.55.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "e8bef9221fb858dd3176c6ec330b240d" + }, + { + "dataPath": "params_shard_313.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.56.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "2eda77ce88575466b8ab551d08e69afa" + }, + { + "dataPath": "params_shard_314.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.56.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f094869caf1757b7aaf465ff09df9044" + }, + { + "dataPath": "params_shard_315.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.56.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "bb989db56a3592e21fa331803d577abf" + }, + { + "dataPath": "params_shard_316.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.56.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "73159186480ae8cc0c122063b848bca3" + }, + { + "dataPath": "params_shard_317.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.56.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "3631cc9b260d624631fac0fb7aee487a" + }, + { + "dataPath": "params_shard_318.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.57.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "2b975de5103ab4826c95b1554577a3c3" + }, + { + "dataPath": "params_shard_319.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.57.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "1686e0d8ea6088499a100b1ec7645d8b" + }, + { + "dataPath": "params_shard_320.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.57.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "219ce5d38cbe08c3c2004d06a997884a" + }, + { + "dataPath": "params_shard_321.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.57.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "64cfb78cd01b54971c131ad8d91e5fe3" + }, + { + "dataPath": "params_shard_322.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.57.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "5b6f5b31dda72af688b5b359915f71f1" + }, + { + "dataPath": "params_shard_323.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.55.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.56.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.56.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.56.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.56.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.56.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.57.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.57.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.57.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "61b76a0decb6c0f7636a53a80f53376e" + }, + { + "dataPath": "params_shard_324.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.58.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "adec754cab077ae6286e56eb2c3e4d97" + }, + { + "dataPath": "params_shard_325.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.58.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "d4f1a7991546b0b79da10b5e6b8964b1" + }, + { + "dataPath": "params_shard_326.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.58.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "8e586dac4abc9294855ef6d0667a806b" + }, + { + "dataPath": "params_shard_327.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.58.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5a8561f77c83adb62d295970a0840f11" + }, + { + "dataPath": "params_shard_328.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.58.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "f295672cec55724618d4916f07d33ae3" + }, + { + "dataPath": "params_shard_329.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.59.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "8f05d660c3b4a94b02226ab62630554c" + }, + { + "dataPath": "params_shard_330.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.57.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.57.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.58.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.58.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.58.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.58.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.58.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.59.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "c9cb02004ee5341124ae5b049d04f7c8" + }, + { + "dataPath": "params_shard_331.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.59.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "f1029693d0ae4d969fdd255ca06f7334" + }, + { + "dataPath": "params_shard_332.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.59.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "5710ee296fd956cb5585316b2929d19f" + }, + { + "dataPath": "params_shard_333.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.59.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5fd9f53a93308d81924f30f2c7f35a13" + }, + { + "dataPath": "params_shard_334.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.59.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a46ed7ab1ea038c75971f810a3207b07" + }, + { + "dataPath": "params_shard_335.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.60.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5d5039c5c6c02d1239a2ab4e7ac3d5c5" + }, + { + "dataPath": "params_shard_336.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.60.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "08203c7da48170eab3947128f3e0e7ec" + }, + { + "dataPath": "params_shard_337.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "7bf6871a254b372852278cb170cc7152" + }, + { + "dataPath": "params_shard_338.bin", + "format": "raw-shard", + "nbytes": 26927104, + "records": [ + { + "name": "model.layers.59.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.59.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.59.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.59.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.60.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.60.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26894336 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26910720 + } + ], + "md5sum": "a1bf89d18b64aa841a42e755ec3f2e97" + }, + { + "dataPath": "params_shard_339.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "91d9f0ae6c6f51857cde4a4495f71dfb" + }, + { + "dataPath": "params_shard_340.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "6f285f929a116a2202575d5cc6700358" + }, + { + "dataPath": "params_shard_341.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.7.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "beb5be16857008e54581947ded779993" + }, + { + "dataPath": "params_shard_342.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "933b484ec4ada993895d66ca0f61f9b6" + }, + { + "dataPath": "params_shard_343.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "d0c361b2b27bae042ac9795645ada9d7" + }, + { + "dataPath": "params_shard_344.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "ee5f8ad4e04dfdce19ea99ee80eb5f39" + }, + { + "dataPath": "params_shard_345.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "adbdf11672f0641f5c04b3554cee9594" + }, + { + "dataPath": "params_shard_346.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.8.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "43d074b78ce7c4385a3f3459646c664f" + }, + { + "dataPath": "params_shard_347.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "7d436c7ac268d4ffd1d6859e21ac3224" + }, + { + "dataPath": "params_shard_348.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.8.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "e7e1fb01d3f7ab5846faae39ba74eea5" + }, + { + "dataPath": "params_shard_349.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.9.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "d06204c9526f6f38552206e99f38fb1c" + }, + { + "dataPath": "params_shard_350.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "f5ecfe8dec949d980af8f6d635dd9030" + }, + { + "dataPath": "params_shard_351.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.60.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "99fcf0b09696360101c04a2f91827d6c" + }, + { + "dataPath": "params_shard_352.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.60.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "21663188e7cef19c1955a053d82b0006" + }, + { + "dataPath": "params_shard_353.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.60.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "7fa74f516b691dc1d76552942d810b84" + }, + { + "dataPath": "params_shard_354.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.61.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "2e6f999a19a0ce80feeb1f4b1ac0b242" + }, + { + "dataPath": "params_shard_355.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 3358720 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 7557120 + }, + { + "name": "model.layers.60.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 10915840 + }, + { + "name": "model.layers.60.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 10932224 + }, + { + "name": "model.layers.60.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22679552 + }, + { + "name": "model.layers.61.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "0ecbe582fc947c200a6c0826d0779f03" + }, + { + "dataPath": "params_shard_356.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.61.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "85ae93c087209d6ac7ef2ac93e6379d0" + }, + { + "dataPath": "params_shard_357.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.61.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "aec491c3420fc98d04da90a7485a642a" + }, + { + "dataPath": "params_shard_358.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.61.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ca09d034d1d5bfb194db40dd07629463" + }, + { + "dataPath": "params_shard_359.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.61.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "9e350b0e6e2cc107bc5816f989c875c8" + }, + { + "dataPath": "params_shard_360.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.62.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "10c8b6e51b8fb3e51806dbabd09fea5e" + }, + { + "dataPath": "params_shard_361.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.62.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "8a118d6fdb98f111907f278d5f3c54ff" + }, + { + "dataPath": "params_shard_362.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.62.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "774ff63b3009edb93f5d0399e8a06451" + }, + { + "dataPath": "params_shard_363.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.62.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0b5316d2273097eb3d8faebadf6e689f" + }, + { + "dataPath": "params_shard_364.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.61.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.61.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.61.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.61.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.62.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19320832 + } + ], + "md5sum": "30e118e6ceeecb13a463a9b681c47c3d" + }, + { + "dataPath": "params_shard_365.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.62.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.62.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "f97d9ed3e6962e2f0086408b31b8a5d8" + }, + { + "dataPath": "params_shard_366.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.63.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "7aeb285b3d1173940179020c1df88cea" + }, + { + "dataPath": "params_shard_367.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.63.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "d5c66a99869ae17b2ed1f546b9e24b8a" + }, + { + "dataPath": "params_shard_368.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.63.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "69eeb54fa41c38829dc6f938b859cd60" + }, + { + "dataPath": "params_shard_369.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.63.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "2911e16971a41fe6257cec4a9f79ceb3" + }, + { + "dataPath": "params_shard_370.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.63.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "5cffaaa510259e2b4c86177e02b14ea4" + }, + { + "dataPath": "params_shard_371.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.64.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "9c06bf5c650937b2e83f6bce03b0f50a" + }, + { + "dataPath": "params_shard_372.bin", + "format": "raw-shard", + "nbytes": 22745088, + "records": [ + { + "name": "model.layers.62.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.62.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.62.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3375104 + }, + { + "name": "model.layers.63.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3391488 + }, + { + "name": "model.layers.63.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3407872 + }, + { + "name": "model.layers.63.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15155200 + }, + { + "name": "model.layers.63.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15171584 + }, + { + "name": "model.layers.63.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19369984 + }, + { + "name": "model.layers.64.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22728704 + } + ], + "md5sum": "e464e2ddbc8635c39c6887ec06e7fed8" + }, + { + "dataPath": "params_shard_373.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.64.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "7e1613a5a0a0a2be9baf614e37e7a6fc" + }, + { + "dataPath": "params_shard_374.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.64.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "0b09e09d8651f3dfd82344c9488f41bf" + }, + { + "dataPath": "params_shard_375.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.64.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "2d75b501cbd8a9a4ceb27511f545a07b" + }, + { + "dataPath": "params_shard_376.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.64.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "f3d4f77ed1df93b975c0695d303f6cfa" + }, + { + "dataPath": "params_shard_377.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.65.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "66fd5ec9305838ee44aa0d58593373a0" + }, + { + "dataPath": "params_shard_378.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.65.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "95f7ef21f996c0557c7c812324ab7f1e" + }, + { + "dataPath": "params_shard_379.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.65.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "5a15d4eb02212099df9c6d7b933ed653" + }, + { + "dataPath": "params_shard_380.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.65.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "9edcbd93b9c0744631d65a3594ce603c" + }, + { + "dataPath": "params_shard_381.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.65.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "754ee6c92eb5244121a59577d250a8f2" + }, + { + "dataPath": "params_shard_382.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.64.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.64.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.64.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.64.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.65.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.65.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.65.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "712fbd51d51918eb3562cc167eeeb373" + }, + { + "dataPath": "params_shard_383.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.66.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "e6cde066167f770f29b15db824b0bbad" + }, + { + "dataPath": "params_shard_384.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.66.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "e9d53950c470d4caf448694fffa4161a" + }, + { + "dataPath": "params_shard_385.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.66.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "fe830cb45e2e8614540ae4e2289a9717" + }, + { + "dataPath": "params_shard_386.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.66.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "ffb63336fa9bcf26aea5a7851af39c15" + }, + { + "dataPath": "params_shard_387.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.66.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "387ac4f621ef2d23dcc59f8ffb1a8db0" + }, + { + "dataPath": "params_shard_388.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.67.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "80a9f18a1afe49db0d45c2271d50b91c" + }, + { + "dataPath": "params_shard_389.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.65.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.65.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.66.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.66.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.66.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.66.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.66.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.67.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "0546a50850d490ab1f3d74f8f2539978" + }, + { + "dataPath": "params_shard_390.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.67.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "dfa260ab140f5297c369ad0716d7a073" + }, + { + "dataPath": "params_shard_391.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.67.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "16b91714b5191f021d67b3a69fa2b4b6" + }, + { + "dataPath": "params_shard_392.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.67.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "9626f2615856f9519c6efd95df30a7b5" + }, + { + "dataPath": "params_shard_393.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.67.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "032e48f68af5b2694c938ee54a47c913" + }, + { + "dataPath": "params_shard_394.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.68.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "eca3589d1fee9a402a719caae5bba24c" + }, + { + "dataPath": "params_shard_395.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.68.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "66621015fc14e5d27247346879bde8f7" + }, + { + "dataPath": "params_shard_396.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.68.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "1047d4414bc92e7cecb7c37d78983e09" + }, + { + "dataPath": "params_shard_397.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.67.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.67.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.67.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.67.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.68.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.68.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.68.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "8751867dc7e85692692082cab7832dd1" + }, + { + "dataPath": "params_shard_398.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.68.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "511bbf7409a6cf864469437d0a8f25cb" + }, + { + "dataPath": "params_shard_399.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.68.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "ecab4f347397067d090d612552e870fb" + }, + { + "dataPath": "params_shard_400.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.69.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "872dd05f65726d3020dc8aaa9972e0e5" + }, + { + "dataPath": "params_shard_401.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.69.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "01ad9626a76941d09ee0e90298e2c58d" + }, + { + "dataPath": "params_shard_402.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.69.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "217653fa8db45f3863839bc4f2a481fc" + }, + { + "dataPath": "params_shard_403.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.69.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "36ee9e883f50d4f0b9d170381edbfccb" + }, + { + "dataPath": "params_shard_404.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.69.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "9a7c65653b3ab679e03ae08d68d0173c" + }, + { + "dataPath": "params_shard_405.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.70.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "b0abfa9eb557bc878932ce02b59e5d78" + }, + { + "dataPath": "params_shard_406.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.68.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.68.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.69.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.69.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.69.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.69.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.69.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "321e52fb7f11787d0f6f89bb0f2df8bf" + }, + { + "dataPath": "params_shard_407.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.70.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "9f1da66df2c9a4a90a96235ef60aca81" + }, + { + "dataPath": "params_shard_408.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.70.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "71a01680a8ffffd06162a89d5fd1126a" + }, + { + "dataPath": "params_shard_409.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.70.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "4f9ed8340a0cbe290e33ae6d3b229053" + }, + { + "dataPath": "params_shard_410.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.70.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "1d994ae3402d82dcd9923d76a8a3220c" + }, + { + "dataPath": "params_shard_411.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.71.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "00f6d895638aacc485bb71489b2cee65" + }, + { + "dataPath": "params_shard_412.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.71.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "548852fb8bbdf7abb7ac72bf46525b43" + }, + { + "dataPath": "params_shard_413.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.71.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "a93abbd7030cf66d6282e13a54317f0a" + }, + { + "dataPath": "params_shard_414.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.71.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "793c8f4c45b69ec34ab5d64e5ea92d44" + }, + { + "dataPath": "params_shard_415.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.70.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.70.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.70.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + }, + { + "name": "model.layers.70.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19304448 + }, + { + "name": "model.layers.70.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.71.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19337216 + }, + { + "name": "model.layers.71.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19353600 + }, + { + "name": "model.layers.71.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "0ca90d8ad7b4b25ff0f2d842dfee0a21" + }, + { + "dataPath": "params_shard_416.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.71.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.71.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "6bf1c45cf1b2319a31784a3dee8df616" + }, + { + "dataPath": "params_shard_417.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.72.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "e4752e0a8eda01cdf524bf0c41896c95" + }, + { + "dataPath": "params_shard_418.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.72.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "22bb56583bb30bc7f5e227ce6fb38dc1" + }, + { + "dataPath": "params_shard_419.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.72.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "630d0b1b79fc8a7bd9496b0cc0ab513c" + }, + { + "dataPath": "params_shard_420.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.72.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "e05c0c1623af7256b4c25d88083c5096" + }, + { + "dataPath": "params_shard_421.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.72.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "53118650ca8e6dad490bbee54a116c4a" + }, + { + "dataPath": "params_shard_422.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.73.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "8c688875159a8749a98cea5b0aef13f5" + }, + { + "dataPath": "params_shard_423.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.73.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "a0340ab10cb1b353a40bc8e07e9a1dfa" + }, + { + "dataPath": "params_shard_424.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.73.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "0959570afe065aa4e8fda29dc334ce91" + }, + { + "dataPath": "params_shard_425.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.73.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "8197ffaa7f80701ecd038617e67b4fe7" + }, + { + "dataPath": "params_shard_426.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.73.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "cf54534c0fd052dd395d651c11888de3" + }, + { + "dataPath": "params_shard_427.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.71.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.72.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.72.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.72.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.72.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.72.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.73.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.73.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.73.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "674878ba5cafbd72b88b409b638ed8ae" + }, + { + "dataPath": "params_shard_428.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.74.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "3af16f0c7cecde6ae716930a6a56524a" + }, + { + "dataPath": "params_shard_429.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.74.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "cc87b1ae51400afe734f2f99154bb0f4" + }, + { + "dataPath": "params_shard_430.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.74.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "9d7997fb59a2673b75abab20ab165d0a" + }, + { + "dataPath": "params_shard_431.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.74.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "9760f81e689e7b1c4e7abec946fd87fa" + }, + { + "dataPath": "params_shard_432.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.74.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "9cb21a3105d7c0e02c84ec6de23a8a85" + }, + { + "dataPath": "params_shard_433.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.75.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "cde6011b0148cf8373d00f0c4b402482" + }, + { + "dataPath": "params_shard_434.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.73.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.73.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.74.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.74.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.74.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.74.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.74.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.75.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "55b82ec280c7e709b8bad227ea7e192d" + }, + { + "dataPath": "params_shard_435.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.75.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "d2c7b0913bec38b15593ec468dac6e46" + }, + { + "dataPath": "params_shard_436.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.75.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "2352704dd6d0b92d9a854d47f73a6757" + }, + { + "dataPath": "params_shard_437.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.75.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "4006990df87d891b1b4ca614a72fdffe" + }, + { + "dataPath": "params_shard_438.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.75.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "614def14bd669328a6c6be3cebcc0b49" + }, + { + "dataPath": "params_shard_439.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.76.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "3733ddddbd61a8f4fd1e734cf131c800" + }, + { + "dataPath": "params_shard_440.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.76.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "f2aa893d5127b422748b34d9d6f4e96e" + }, + { + "dataPath": "params_shard_441.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.76.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "bd3696e959d11fb0633341782dc99e22" + }, + { + "dataPath": "params_shard_442.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.75.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.75.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.75.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.75.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.76.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.76.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.76.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "ec8ad5c0a7e24efd06a8523bd489caed" + }, + { + "dataPath": "params_shard_443.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.76.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "7fbfbe3a00f217cea9d88c42134493a8" + }, + { + "dataPath": "params_shard_444.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.76.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "f22d82d2a32c1bdae70adb28cfe9f3ca" + }, + { + "dataPath": "params_shard_445.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.77.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "e8b11301c86cb038a96e517c3827e60e" + }, + { + "dataPath": "params_shard_446.bin", + "format": "raw-shard", + "nbytes": 188088320, + "records": [ + { + "name": "model.layers.77.mlp.gate_up_proj.q_weight", + "shape": [ + 57344, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 188088320, + "byteOffset": 0 + } + ], + "md5sum": "98b07c88aa2fd4aa156e61165c551587" + }, + { + "dataPath": "params_shard_447.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.77.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "dc1df9f45b9c4069ba565691a191f506" + }, + { + "dataPath": "params_shard_448.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.77.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "e9a4a37b87edfa911f22f7202719eea5" + }, + { + "dataPath": "params_shard_449.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.77.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "e3ff1add024f4d9990a871e77de7c564" + }, + { + "dataPath": "params_shard_450.bin", + "format": "raw-shard", + "nbytes": 93978624, + "records": [ + { + "name": "model.layers.78.mlp.down_proj.q_weight", + "shape": [ + 8192, + 2868 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 93978624, + "byteOffset": 0 + } + ], + "md5sum": "a4fa127b4165456e1fb5a994babd79ae" + }, + { + "dataPath": "params_shard_451.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.76.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.76.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.77.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.77.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.77.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.77.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.77.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + } + ], + "md5sum": "98eb335baf3035df9ca7579756fdf350" + }, + { + "dataPath": "params_shard_452.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "model.layers.78.self_attn.qkv_proj.q_weight", + "shape": [ + 10240, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 33587200, + "byteOffset": 0 + } + ], + "md5sum": "bc233e4afdc1e5047d38be2a1ff58c87" + }, + { + "dataPath": "params_shard_453.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.78.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "2d496db30e5f337d98fd9f5a45514e68" + }, + { + "dataPath": "params_shard_454.bin", + "format": "raw-shard", + "nbytes": 19304448, + "records": [ + { + "name": "model.layers.78.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.78.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11747328 + }, + { + "name": "model.layers.78.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15945728 + } + ], + "md5sum": "195728e28567cac968a45830775112ca" + } + ] +} \ No newline at end of file