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": "d54a62e9cab21ddc05182e07e2786e0c" + }, + { + "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": "395d9caf2bb0d5973a584b903ba5d04b" + }, + { + "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": "ff3f529a8bbe11c1aa3b393e8da29e5b" + }, + { + "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": "8c6f4d67864bbc1bf600ad1ae5b3ea5c" + }, + { + "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": "b0d600a479894cd42b0529bab0a03df4" + }, + { + "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": "2bf10be88ca61c2e78d91de4e562f3a8" + }, + { + "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": "a00f7a03d1ba55099dec2fec43c01bd4" + }, + { + "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": "6c5bbc14ad0a46e2c8316b2801eb7294" + }, + { + "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": "5501dff64322180d467c20037d4244ab" + }, + { + "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": "58381ad0023f210935b3ad4c6f8e7b35" + }, + { + "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": "70be19ba6f0f3144bc37fb51c4522c45" + }, + { + "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": "0e923cecc32bbe5fabc4c66a67fce5f7" + }, + { + "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": "ae80ac9eabded000ad9922724cb41609" + }, + { + "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": "7fce7b2aaceb9a540a8ee1430c7f3ba4" + }, + { + "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": "477957bfe3dd4afd6924392cd7c854ea" + }, + { + "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": "29feb063350a34afc5b720d18d52b075" + }, + { + "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": "b639a134d1115f8148541b69e14cd31b" + }, + { + "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": "68b8fcd6db03718cfb3ecfcf44e15acc" + }, + { + "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": "9985781aefb1b8ee2b1f2b29f480ba7c" + }, + { + "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": "772c83776dfa54b6a6f9a97e529855da" + }, + { + "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": "dbbd1d69758002fdd901af1492e61629" + }, + { + "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": "8468f881a962408d94d81072e3f8d6bb" + }, + { + "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": "53699a08416db230ecd09ccef3c70c7f" + }, + { + "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": "059c6fada6395aae162732425028a916" + }, + { + "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": "9f28a036bdb2e7f83e77a00234a262d9" + }, + { + "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": "c784d8261494e758b0aa4a68ac75e04f" + }, + { + "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": "ba37f314d2a9b3c64ea47c45cc3eb57a" + }, + { + "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": "183fc514ef46a424ebdc4f97c937dd22" + }, + { + "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": "543117a8b009e43e7afbf61cbb562088" + }, + { + "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": "49148dde9a81999a8a5bc9191cd588e6" + }, + { + "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": "045e6458263f9ce6aaf9de24f19f64ac" + }, + { + "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": "13775d67866ef1560e4f3b0c3e84ccaa" + }, + { + "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": "3251104b193fc2d0c1cf3f3011750778" + }, + { + "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": "9fe884085d95990d2f08599ccc262167" + }, + { + "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": "896dadf2be68ad3ea128d8debfc2e2d2" + }, + { + "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": "1c2213dc32b6d730adfe6aeebf97d8ee" + }, + { + "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": "1d254be490572d32d65c4eb095907d65" + }, + { + "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": "e91a637a6cbc4ec1978dafc682b34abc" + }, + { + "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": "1f60fa6cafab00ea46cbfc94e404acf0" + }, + { + "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": "0ce7e7036346af5c4b25a90f75f35cd0" + }, + { + "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": "1fe62af5fd77464964d93ae6ba9fd015" + }, + { + "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": "727e0e7fe901f1eefc95e16441013310" + }, + { + "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": "f0383ee3721ece83dafc8f67688d1cc1" + }, + { + "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": "e7b8e54e7d296b1616d4a9ab5e267b75" + }, + { + "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": "1c816c55ce0f7f7170efbb9092dc517b" + }, + { + "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": "8a1de9f54703400c28ed6e69b3bcae35" + }, + { + "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": "bb294ace18d145cba2a689ce85fe7a48" + }, + { + "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": "c9b9db73ff52cc2d06528e3ec63a075d" + }, + { + "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": "cbb09ff4f570994bf60936c2420c5d65" + }, + { + "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": "8a3fef073262d304a4b35907831cd66c" + }, + { + "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": "debee6ce0d7982509cd098afea72c1d2" + }, + { + "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": "cc2f85d151caf5ffd2249fe50b9dabbf" + }, + { + "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": "47c130d9b9de9bd0ddbf0e1ac0527f2e" + }, + { + "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": "c6264ad1d1778ff339445fe80f6232e6" + }, + { + "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": "8aa64480107d8b55289271dd2835142d" + }, + { + "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": "923f8dd1bb1b7147939fc04b39177773" + }, + { + "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": "9f706135b56bba2cc21bf92d96b33aba" + }, + { + "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": "45039ea9e63d7da8ed23510901c75827" + }, + { + "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": "9f651e1247a05fc2983cb02b8f4971f5" + }, + { + "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": "cbdc2f4b40236ae5a5ed5f4795720d9a" + }, + { + "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": "c344a5a8b76c78c9f8e3ddb30a0613b5" + }, + { + "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": "94dcf6f75da190e69eeeedcf0289156c" + }, + { + "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": "555d41b504029b27e2834b15e79088c4" + }, + { + "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": "c866f8a984541f7416860f7b0380f8c1" + }, + { + "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": "5c5db0269f29aa6711e9801ccec7281a" + }, + { + "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": "e37a02e7aef781f97b2f2d132d5de283" + }, + { + "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": "6d35ada7607b32b4818911af33e72982" + }, + { + "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": "800c53aa0f1c7122a4dedda2ba8fd0a5" + }, + { + "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": "5998c7c4fcb64d871741af803fc84d70" + }, + { + "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": "3456ce8e95b11f96965c926d18e571fb" + }, + { + "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": "96ea19171675fdd4621a5b2d0a1ce2c8" + }, + { + "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": "17fcbd4d3613eca13e14d347e7a3e3b8" + }, + { + "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": "bfc1cddce73c4da14fdea3f6b9f47fc1" + }, + { + "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": "a5ed985bfbfd9276287d62b917af7d99" + }, + { + "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": "d23f4bc842444f51143ade0956062c32" + }, + { + "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": "1c7130be3bdac725a99684c78987716f" + }, + { + "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": "6b2c4a20b6369acfe9f93473cd1abc55" + }, + { + "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": "6b8b31e03a63c2952b8ee702c49484e3" + }, + { + "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": "24dbdf3b012dd846ea93169ad2309ba0" + }, + { + "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": "372a701d51b251cdfd34d57d91a44598" + }, + { + "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": "bebb13fe61b73983b01928dee833a3a3" + }, + { + "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": "faabb29ab41e37ae892c57693632c055" + }, + { + "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": "63d55cd542d066b504af5a5439d35532" + }, + { + "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": "9f3813ebf596a1431229ba7b759c7e73" + }, + { + "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": "13443ab1c60a25292d2f945a374e5bc7" + }, + { + "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": "3a5ad84d03791214dc27520486e216d2" + }, + { + "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": "24168a34a4f25db538f95c004aefa640" + }, + { + "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": "7a3ca1df263346cce19114e48ee62a60" + }, + { + "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": "d01d650dfa6c19face9030dc9ba25aee" + }, + { + "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": "6298766803575885cf3f7cbe2b265480" + }, + { + "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": "a7356ab9d90d1055e4cfce9d7ea97cae" + }, + { + "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": "50177fac4ee4c596520205a822414aea" + }, + { + "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": "d0d9f11cb77c8ff85828f3d5bd782a22" + }, + { + "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": "eabf1f8f37568a9ecf29e11f898b5a9d" + }, + { + "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": "9ea690931ce61eabd27dc1743fb00444" + }, + { + "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": "9eb1d28ff8fb0c9220e6b5665d96844e" + }, + { + "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": "31282f37d6e6c983463d4d7e57a328ba" + }, + { + "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": "2fe2cd9472e9514917e87f458c74fbbb" + }, + { + "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": "70ebb4d14e7f6d9061ae2d4f0bd2ebfb" + }, + { + "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": "bf8e87906cbf4ba6fe2b7a0bc3787ec6" + }, + { + "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": "3af01fdb156f8fb313a1617deabf96d2" + }, + { + "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": "916e1c720bd2dc611a6f6b39e1ae253a" + }, + { + "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": "075cef8c3f1289447d7b54e11f0a0494" + }, + { + "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": "c3705adbbebdd76054698a53fa9449ca" + }, + { + "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": "0d7dbca1751b6b3b391c6b9cd84dc0c2" + }, + { + "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": "9ecf8eefbcde5c17584baf8821eab1cc" + }, + { + "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": "ad85e8ca3d41c7197e20b310596936d6" + }, + { + "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": "941eb3118de69347cf3e43304492a044" + }, + { + "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": "10ed72e3f63ca406eb76d0135e4b6198" + }, + { + "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": "1091738f536716d57012194dac1a4dfa" + }, + { + "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": "7ff9e5181b5c90e15bf579a110d93936" + }, + { + "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": "8b4aa9a8dfaced8d8384fd0b3cc47df6" + }, + { + "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": "b412e678a0618fb9562e1b4ddab412d2" + }, + { + "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": "d76c7ccc88313c94fe79340b6fc1d7fb" + }, + { + "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": "7fee9f06ee9c4fc4915c18531f1a1847" + }, + { + "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": "406b60542a12d170f37edcbd81052b76" + }, + { + "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": "a0d254f320dae62360f566c559c0e4cc" + }, + { + "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": "a1569b70b850e8a33b5663486072d73a" + }, + { + "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": "ff8bae7b5593874f0d2dddfda58a412c" + }, + { + "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": "fe704b9f6ae36ff4e950c05e00ad194a" + }, + { + "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": "7098b10b3cf82f7f975ce9815b2995bd" + }, + { + "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": "dc1940a5692fda5bd7018d876cf0a136" + }, + { + "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": "14c446581193398e230816e28a5053d5" + }, + { + "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": "c5a4d607feb1435ad6a62ae2aaf93bf3" + }, + { + "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": "895f3611c7380de80e45b7f132fa478e" + }, + { + "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": "a89dd883934908af97e08e1a2710bc12" + }, + { + "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": "16fa5a99732eb464535e1fff9b972855" + }, + { + "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": "23060fa859787dafd297245b0a88d338" + }, + { + "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": "3e877876e2882409854da94131f5124f" + }, + { + "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": "2a49e8a6a56ff30d48f9c0497810f52d" + }, + { + "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": "116d8d0a8ee87078e64d5269c8e8b5cc" + }, + { + "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": "90aa5d29eab0792047b56bc7210fd61f" + }, + { + "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": "300bdc1aa30f4bd59acf32a97b09742d" + }, + { + "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": "d823142a135bb2055fe3de15cce00189" + }, + { + "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": "39aebba9a7c7197ac712f5da4ce3ed95" + }, + { + "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": "f6b12e1c8eb73d30a2a4e04c778e952f" + }, + { + "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": "cfff1a63ac021b90bbd3ef33282953c0" + }, + { + "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": "74df906c809a660043839ac71af4bfe3" + }, + { + "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": "ad70c4c88591d0e722c37ee14a5082af" + }, + { + "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": "f19614c1ae428e657973cf31374c8188" + }, + { + "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": "10d5b100d518658cee8d7f0ef76b3312" + }, + { + "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": "624594c2e8b2d9d02f32a759dd2304a9" + }, + { + "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": "63b5655d8c9e2a7fde38d38072c473f2" + }, + { + "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": "0738c7a3026fb6e637c104c0f4e74d05" + }, + { + "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": "e8d5bb1c9b3b75dfcece99d0c70d5af3" + }, + { + "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": "ed241f7c2fd980f05c57fef496bc53b6" + }, + { + "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": "0dbd74c13f476b0fa02e2cdca4a35cf5" + }, + { + "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": "71c52adf6c6131625b4c14b3bf285575" + }, + { + "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": "a0b4693a36f2bb8e5acc4c65df9b8a95" + }, + { + "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": "288a9f66ed8d2a96887cafae76f92e81" + }, + { + "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": "c550e69b0f855d36beb2aac2a362f443" + }, + { + "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": "374953ab0c77a73820a7b49ad6140d71" + }, + { + "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": "40c815a6d724a87e527f56be533ac64a" + }, + { + "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": "984268a61f739fce0192a306c556d172" + }, + { + "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": "6fa12cfd0ee7576ae379b73579ce1363" + }, + { + "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": "4e295fe9535142fbd4bc146f5f82b48e" + }, + { + "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": "72d40a774a8b55cdd2916b6fb2e1994b" + }, + { + "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": "03449d6ed9410bb1dd7c30bc75d9fa52" + }, + { + "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": "f9d0570096e1c0698964db1c45b55b72" + }, + { + "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": "2fdd3ba9a5c8f70d62f13186d7c8803f" + }, + { + "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": "ce6afe286907a53fc420e7ca64b41826" + }, + { + "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": "23ec26dcd64ab0ccbbda2858176c9662" + }, + { + "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": "911af899002876cb89aa50e220c317fc" + }, + { + "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": "7955cd79ab3e8362ec15fa68b39f1806" + }, + { + "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": "0c07c7c05cec72cdee3b2c6420f682b0" + }, + { + "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": "90de00ae7c36e5635df8ec32c8f16df4" + }, + { + "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": "9ddd74734cd813b713d978612cc69177" + }, + { + "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": "526abb9885689ef2434fce560ddad9cd" + }, + { + "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": "7530fa7d6e17a399ce6e4a834f1e53d3" + }, + { + "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": "a93757d88344de77e01cd9e5703e3dda" + }, + { + "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": "de4038f7287cfe1e47fbdef181eb1bbe" + }, + { + "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": "8fa10cc1b293da3d7783f47dddf4cfdf" + }, + { + "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": "f5319097e95f8c96a6199cdb56d66767" + }, + { + "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": "63003b043faf3fbc35e4b50b35fd7027" + }, + { + "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": "4f9d711992a5a692d1e66112321cb1ed" + }, + { + "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": "9def2286b82c6f67105026a64068448c" + }, + { + "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": "4cb72c18fdf6f2b3630f0e5ba34e42d8" + }, + { + "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": "c8c4c601127800c63317a74c6dbffba2" + }, + { + "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": "374746bc5105231c7f3fa93f44ff6ba7" + }, + { + "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": "fca8829c2fb9463b1b0cd0db052fd1a4" + }, + { + "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": "f409ef7778e0286082cf2a8f592b17c6" + }, + { + "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": "654f623db5824706c7b565bf8afd3755" + }, + { + "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": "035ed28593d294fa252813e102c89f8d" + }, + { + "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": "a0b3ef979af1e3a4c44ab0cfc430195c" + }, + { + "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": "efc3ffc6588aeb5fe3da9c6991302701" + }, + { + "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": "3a966285289ef0388b8d6ed87a1553b9" + }, + { + "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": "86e0822b21a813bfc41b21d220f3a7d0" + }, + { + "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": "608f07f6fbb28b5d8eca199995bf9b42" + }, + { + "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": "14cbd389c7db4ef928e28384808f08cf" + }, + { + "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": "3e8dbfa99ff397ba9ea0e2ea39488f02" + }, + { + "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": "943894e63724178fe26a57a7f2ce8ea0" + }, + { + "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": "052cac5831a5db7ee26fb2258ffc8bd9" + }, + { + "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": "d98291ad658db9b5acf81b129267dec9" + }, + { + "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": "0db37e5c06a3bd536c237a8eec7901d4" + }, + { + "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": "7b9ea9d35c8841764d4fda856fd4f487" + }, + { + "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": "80dd69da7e04a9cc9fd5f6228df74a24" + }, + { + "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": "f0f3b7e8fff0c47343dc210ad2694d27" + }, + { + "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": "f37b25ab40cc7c27610d3fda2518d1d7" + }, + { + "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": "c9bb013f7a25876b8ddf2cb285c4b7c1" + }, + { + "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": "fbbbe64abb9ac903c545f268902240df" + }, + { + "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": "c5ddbe456cfb8e3cf7bf93a2fd5ce714" + }, + { + "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": "77b1b1ce014ec1887756d33488d90c58" + }, + { + "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": "e2c948d06f98a75ba91dc8b163915dae" + }, + { + "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": "6261aed539e3a59da6c7a3d4903ea032" + }, + { + "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": "a0fe6a3f6be8074ef74560eaff92da2b" + }, + { + "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": "a6db1247982c9621ec7d79c079a8ad1f" + }, + { + "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": "73b1a4862a7352ddaeceb8a2aef0919d" + }, + { + "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": "b4e27f702c36255068544fd85ee770d5" + }, + { + "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": "6677ba3c99d54a76e25ab78854cdf22f" + }, + { + "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": "94112b9d141fd921c4f58df9405f04f7" + }, + { + "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": "2a09746f9c99503084bbd1e6b125d1ed" + }, + { + "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": "140af1644075f3c9f2bebd6709b819cc" + }, + { + "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": "daac6fe6cdb4f7f5ae38bc81493ceaff" + }, + { + "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": "eaed9053f2f933cf5b4fb92d1268825f" + }, + { + "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": "aa1c3d7a0dad8ba776f2f1ce73372dae" + }, + { + "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": "b7a657cccce0b58c27068d77bf9372a6" + }, + { + "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": "70dc849d5aaafcf82a5f223fc41a44bf" + }, + { + "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": "6ed0cafe50f6b67e081fc3f6ee01a244" + }, + { + "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": "9e476b43ce330359db33d546506844ee" + }, + { + "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": "08e1f6d95457cd376db8529ad87bdf69" + }, + { + "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": "6a60c2ed09a74fc4ea1d87735e487ec9" + }, + { + "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": "608511594e311e411905cc2e93c7aa6a" + }, + { + "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": "46de445936c4820a1e304443e238b7e0" + }, + { + "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": "a4236d62f7a7af76b6432650fb3216ad" + }, + { + "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": "4c82bc583262266778099ce86cf904d3" + }, + { + "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": "70fd1a4aec6c2534dc2502acf32e8c6a" + }, + { + "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": "9bca567ef3736eab30d5ae64cae33a94" + }, + { + "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": "6d32b787010ced0fd24aba21b30f7fce" + }, + { + "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": "d4a7f5f7380baf93210bad3825c9091d" + }, + { + "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": "dbef8d16523ad01d4dea73bc99bf1795" + }, + { + "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": "dfea6996890f96432e3356e7e2a70d1f" + }, + { + "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": "5c0ba0a26c750b81406fec07ca3fe583" + }, + { + "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": "e0df1c2a618f75db87e8201d548a874d" + }, + { + "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": "b7a4da264d3ab1c0ef83fc04f7da6c4f" + }, + { + "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": "914395e29b6c173d0ef3c15f1a983761" + }, + { + "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": "d253eff959a4318fce6ce2e2db19c44f" + }, + { + "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": "aea3810f2080f162da2f3fce05164829" + }, + { + "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": "9dd6b0dc7fbaa4bd2d31000e6e747e47" + }, + { + "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": "6d9cd8485e5ad942d46cc1fd5288ba5f" + }, + { + "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": "f53b83a7bf349f846eb9d04134b795ae" + }, + { + "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": "9945a30f42dab71760552def5d4f8424" + }, + { + "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": "f6dbd5c0aed9c7569e1a965d43194985" + }, + { + "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": "2df6d8933ebb4771f2240150bb9371a3" + }, + { + "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": "c68220e28908e7c92614e216754aa2ef" + }, + { + "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": "afe273289b43b5d74653f7f635d275a5" + }, + { + "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": "83811da946a61727bb1a30199778bfb7" + }, + { + "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": "4fdf33ab6eb078764439c9c55e54be5a" + }, + { + "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": "28b2913582e6cc5dd6a3223d3d2bc2c6" + }, + { + "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": "9c34d6fb14b1a0874e46d19a34872021" + }, + { + "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": "b0f05af4d20efb91e3aaa404cbbdc5d1" + }, + { + "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": "467439c64ea50e0aed257940fcedf180" + }, + { + "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": "19e3dc9d692e0cb81433f7a8ea77ba3e" + }, + { + "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": "5bec5725e844924251cae58a0bae8d07" + }, + { + "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": "a301f551b9a17f83f074e911881591ef" + }, + { + "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": "884aba1d6c09f549763b672c1d68225e" + }, + { + "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": "d3c2ba0605309e325da56883d8ffcff8" + }, + { + "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": "a338b3ad38e1c20858e66019363a2d57" + }, + { + "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": "192fc757a9b406854f3f306ff1ab0dca" + }, + { + "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": "799dd5d04234507801f4fd3fbb58b457" + }, + { + "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": "0afbf5964578bb7fbf2183438593609d" + }, + { + "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": "cc04405748b7b2de2785e956165fdf1f" + }, + { + "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": "b01c362f7f6d570a3dd647c8bbb89a31" + }, + { + "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": "d02d185b741a7960fced892533c9f7c7" + }, + { + "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": "43298336f055bd26ab1602dbe4692a26" + }, + { + "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": "0ba6b3b4c3fbd0491682c24a48bbd1a9" + }, + { + "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": "22fde47255c3f5977e06147d0ab35660" + }, + { + "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": "5d84379eb88fbdfd5dfc4117b7605c47" + }, + { + "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": "b8a7b9f8119d4badb593db6be906f431" + }, + { + "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": "9c64cce4eb1c74b8ce4e8c067d48292f" + }, + { + "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": "4f51c7443c5e59addfcb12b1cedfc758" + }, + { + "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": "46b047dd894393bfe3ebc7975a8f6300" + }, + { + "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": "2bf780f2f4cb33cdc3f6c6710ab4f903" + }, + { + "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": "9685b8c6b70b4b8fb8ef03390cbab655" + }, + { + "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": "8ce136332b8b005371dbeb769d31102b" + }, + { + "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": "8154a1737f37b763d19a1e60c418e5ff" + }, + { + "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": "f55bfabe23581f94c5de9c33f1522a2a" + }, + { + "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": "acdb51a29bf66483dde9781dfd7b85a0" + }, + { + "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": "11ad27b2185ad02f415feff8e6303fa8" + }, + { + "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": "1cdb1412e2ce4c7b8bba815c03eaff77" + }, + { + "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": "a9ad990c9a52e4c00a7d3662a9a5d06c" + }, + { + "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": "5a4b7c6089b5b111be03caa6ea4677b3" + }, + { + "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": "f8e274dbe29e974b2e9cf51ff8a735fa" + }, + { + "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": "aa892d3765b6ebfe8e9e2cae20d2812a" + }, + { + "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": "7e3cb69bfd83f38640c1eae677a22426" + }, + { + "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": "b9b7f69a5ae65657206863cc8fbc4d9f" + }, + { + "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": "0ec0451b875b27e4ce8a0708ca6691d9" + }, + { + "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": "20162e938ba0ddfe37c4597c1446c591" + }, + { + "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": "e1bbe928661cd2d9562e0268e268380d" + }, + { + "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": "f2046e1d2adb2b3267c2be353c13a552" + }, + { + "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": "22fa10f67dddc0b24418a395d533cb72" + }, + { + "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": "4c58fdb1fefa97e4984870eeafd17e9e" + }, + { + "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": "d44381a1f45fba48ca93e8dc43817db0" + }, + { + "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": "77e6339aa17c3cc637270f9295ac70f2" + }, + { + "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": "b10f6cd38681d2a827498930e00ab029" + }, + { + "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": "73a0b12e439c60148bda2ecca41fba13" + }, + { + "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": "0f28775556181278ff2e5af27d7eebc3" + }, + { + "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": "06fb7d329287f0ee0a3a5d8cb83ce2a3" + }, + { + "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": "f4b0fe3b05d1292a687285162d2514b5" + }, + { + "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": "f9eb170211227fc9a1774e77104250c6" + }, + { + "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": "c600a9a7cbe197a5f66b0095f52173ff" + }, + { + "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": "d83318ae3142e5e32b3694b7effa455d" + }, + { + "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": "a43f43b5b83738ad4fb29e98d3668a0a" + }, + { + "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": "9970d3a94981ee2102773e43f39583bf" + }, + { + "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": "2a05952e89d570f9c865798923611167" + }, + { + "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": "359b846bf797be212bd510dc4ff138c0" + }, + { + "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": "4eb07dc343e1df4432501278cb0c6ec4" + }, + { + "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": "d75c0ed2c1c256d014e5cfc33241b10c" + }, + { + "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": "af3d27f04a2431bbe94240dea9a43c6f" + }, + { + "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": "78e6661edf7a496561faac43edc47211" + }, + { + "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": "1516eb21f86d6d2b8a2aec1f2e7217f7" + }, + { + "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": "9b041a06f3ec81427c23c1f5cb978cfa" + }, + { + "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": "ed529bad3bda86b0443f8ea7707e2ef6" + }, + { + "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": "eeb9d15676b5e8c3ddd7c4dfac95c818" + }, + { + "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": "320147bea977291a68437cef2ab084c0" + }, + { + "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": "bce55e7da92f4ddfcd1ee38d4e2f993c" + }, + { + "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": "ca091f72562ed17d4637610728d9dd69" + }, + { + "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": "41e91a22ddcc4649349cfcb2bd60c655" + }, + { + "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": "875b88aa4df13253307053fd9baa3b71" + }, + { + "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": "79e764566bb26109634bcce73b088352" + }, + { + "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": "0919aa066ca2688627da76b9a2e76d33" + }, + { + "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": "95e0d1345398eea919c547fed51ccf39" + }, + { + "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": "13f5eba5251994ee8b261bcc7365965f" + }, + { + "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": "e8803123539bf626806cbcc4ad41743e" + }, + { + "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": "9213d54e92751f5f50de1d53c02e49e5" + }, + { + "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": "1d627c01cd227cc78720ab95882ed86b" + }, + { + "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": "effe422fecae602b3523aca16df83391" + }, + { + "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": "410a61d91efec5948d0dd8c6a0a05472" + }, + { + "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": "4e38fda004498b93f9ef411847789831" + }, + { + "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": "a1349e24d3cffd6745f1302736d16a88" + }, + { + "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": "2c9743fa8c2a82b3d9090a2cd77e0c4b" + }, + { + "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": "91fed6ece50f733b9beea3e1827698ba" + }, + { + "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": "ec848631963df6bf1774c314b8b2341f" + }, + { + "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": "e2e646fbe31b359acb4972785297dd1e" + }, + { + "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": "a6e419b22b58af8c63d69ba7cf51a536" + }, + { + "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": "8b79f5984674924c6603cff75a0219e2" + }, + { + "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": "bbfc584475d59de7df745def0b2e0c40" + }, + { + "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": "7f78185427ff9f3c2411d295fa073d88" + }, + { + "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": "d39edb6593184a159ed3818e7660aaa6" + }, + { + "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": "0a36e5c07cb1799319dae7e3ec70fd44" + }, + { + "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": "65d6597ad21ae6a97a5e7450403a6601" + }, + { + "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": "2afa947eda45e33d866bf257832a53bf" + }, + { + "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": "49e7bd6edde5edff7e43cc437e1ffafa" + }, + { + "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": "e544714c3210248085aba2aaa91da118" + }, + { + "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": "058bb8e9eed84f9bc56d66119911a3a5" + }, + { + "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": "812d197dd39b3dfb2bd6faa7b16cad38" + }, + { + "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": "e6da5734d321b723c2e7923f256cc796" + }, + { + "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": "6368fa47de72b8d5aca15bd418143ba4" + }, + { + "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": "e7317b14a9cd2705c42bafdbbe55b4fa" + }, + { + "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": "a14ba0727d6c0234516505864bbb7d69" + }, + { + "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": "fed28e78cb763d062d71eaf0975ffe93" + }, + { + "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": "9ceb7b039afaa52c86569ed5bb42850e" + }, + { + "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": "e9af1ddecc4d2fcf95f01fc92cf59fa8" + }, + { + "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": "79dc540aa2dda7c3fb6a02523c055b2c" + }, + { + "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": "5e3a25f980f03c6a6205281b1947bfde" + }, + { + "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": "bf5548a916b36a9cafb229f9768a41a8" + }, + { + "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": "47781a3124edad6b64fe44ecff298c60" + }, + { + "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": "7b782de40184a86ca1b6ba91a7421b49" + }, + { + "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": "a5784a2394c0f662502412ee7a5531ef" + }, + { + "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": "3dc573c148e0df048e81a49266e23ebd" + }, + { + "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": "6e7819d0c474f844f9f72865e3a9d3e6" + }, + { + "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": "43d758126fea20d27b83f661afd6ff17" + }, + { + "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": "0355100cf6e5f9c9f462823279f24a06" + }, + { + "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": "25b1f83542cfe71b923616ba5ecd6ccc" + }, + { + "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": "cddf81b9e70a953ab71c98e17cd541a4" + }, + { + "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": "194cd19f107f89977b6e056b07247ba8" + }, + { + "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": "3b1037c7c6ccb2d5c39f645f47590ad1" + }, + { + "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": "cbf898721b8b96077a016781be23977c" + }, + { + "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": "e258d9ae43d3f1738e7b1b53ee726998" + }, + { + "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": "04e3b40d32a82975f5901f103811c497" + }, + { + "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": "85813b387026bee903907c69097dc3f6" + }, + { + "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": "7d5647dc050add82c832614aba450a37" + }, + { + "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": "eb095695be98aaf702892ac504ed04b4" + }, + { + "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": "cdf436e69472c1efe739302f5af533b5" + }, + { + "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": "e285877d427542a3223f0fff094a0611" + }, + { + "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": "fb7d6168627d4a0d677b3a1af2ca59b7" + }, + { + "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": "a8ff5c38939c1ca75f104dab192176c3" + }, + { + "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": "511bac2621c89f6790de328d3e4bda40" + }, + { + "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": "604f1f6964c0d035813675d2f8243f9c" + }, + { + "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": "be173a3bbc490f253574cb439d45e4f7" + }, + { + "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": "aa8b15975fcdf9179db1c90fcf635cca" + }, + { + "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": "4c1ff6770c64020e2c3d96e828a3d92a" + }, + { + "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": "25152fccc270abe9fdd33393c36bbb03" + }, + { + "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": "968cb2008b553467e77567c83f869cf0" + }, + { + "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": "c1aa18d04af06573afbeaf1c1e389c73" + }, + { + "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": "436e12eaca43343d3631aba08a7d07ba" + }, + { + "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": "b1e7a8b5f3f2773b13498e2c50196df9" + }, + { + "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": "15cb223ac54da1f53e44421388bc2edb" + }, + { + "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": "0af081eb4c4ad14b3e38bd161ce99739" + }, + { + "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": "8357cbb706b09d47d2fcf06800688fd0" + }, + { + "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": "a7975b912a1c3b94fb0ee84fd2484566" + }, + { + "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": "0908cc7c8d6a3a10c605cab9082fc1a5" + }, + { + "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": "2a60ccafa76ebec17a1eaae3c7f289ee" + }, + { + "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": "1f9ee53cf3be56bd0dde38fb33709f6e" + }, + { + "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": "99218c66b68e2656bf6ee09a2fd663fb" + }, + { + "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": "3f195635e72502fd9252c93b6cb81bd5" + }, + { + "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": "55e5eb44ab73f15f767b6647b23fd126" + }, + { + "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": "bc1a0e2e0e39188c2ffdd1fb62707990" + }, + { + "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": "60c5fb586023808a2a61a4b5152a499d" + }, + { + "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": "15187d9eb29f60c3b2459193101b2506" + }, + { + "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": "e6cff85f1f8977466bcf128c04bc4b2a" + }, + { + "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": "2abfb7f45324410598037f594ccf8f0f" + }, + { + "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": "79eda8f8444f1c2d5a2d427239ab7686" + }, + { + "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": "d3601a44dbf3d75a427d880f98f0f115" + }, + { + "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": "400f481857884b233d9e634a7a04d025" + }, + { + "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": "64702a4d6e707953d47740a0b4b3c758" + }, + { + "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": "08ebea87e8dbe64829d79a885ec7f7f0" + }, + { + "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": "13359b5f259485a6a60bdba98cab2f7a" + }, + { + "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": "58f75534c1a753ce7ac75495ae8a52e1" + }, + { + "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": "4c52fb77aeb1a398e42df4a9a6ec0f5f" + }, + { + "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": "a574212437991b1dad70d5c5635a0981" + }, + { + "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": "cfeea1a7f6e2e03799a65d27669c32ad" + }, + { + "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": "87fcb790797a469d2aff93cef9a79efe" + }, + { + "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": "343788f666c365f3d5725dcbe8dd5521" + }, + { + "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": "7bc512eaeb4fddef53bff008d3d511ec" + }, + { + "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": "807bf85bc70c0d4474f469d480330227" + }, + { + "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": "fde92c7d9433c52aed0d39e82a087e0d" + }, + { + "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": "9e4c17a84da9b163f58b3e8fd9d800ab" + }, + { + "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": "b0d537b1d73818ee00e2773655ad463c" + }, + { + "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": "dbd5c0bb1ea5eac63b5a24c0b3744828" + }, + { + "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": "8f356bb3e20f5debc4ebfded297b3be9" + }, + { + "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": "dce6e55d348420c64141be1d615074ef" + }, + { + "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": "4ac5f657571b636190618c6cc2703e9e" + }, + { + "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": "a10500dcb5043977b4673f5494da014d" + }, + { + "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": "c2a514fa1e171bb7224059aab76637c3" + }, + { + "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": "bcff9f385d4bd34ca7020c42f0310f1c" + }, + { + "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": "491af6d8af9c23ae0825d8dc482f783a" + }, + { + "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": "abd097983e0aa5a7cc68894dfdd15c4f" + }, + { + "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": "84b047ed8eea73037ad591091598ba42" + }, + { + "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": "8e8ed0ee7c7b4880c1814b547239ee28" + }, + { + "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": "fa6a1d3fea315699a65d572812b4cc83" + }, + { + "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": "3793d80c77a73d12cc83eeeee9c04926" + }, + { + "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": "bf06cc235143de15366df98d3113bd46" + }, + { + "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": "fba64d8a7333d061c752984ecc99cdee" + }, + { + "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": "931dd93be1ec01d44ac5de77f7a4360d" + }, + { + "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": "a8ee458d797e61354aa5ffa9625c7c6f" + }, + { + "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": "903f3b4ee73c77101eac09680c977e1e" + }, + { + "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": "5a9dccdf1ec6d39ca4d98b6c0fa3fd45" + }, + { + "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": "863440e5c3aad8a9f50d2d60f1628d63" + }, + { + "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": "b40dc06faac1c1909d29088df526d7e3" + }, + { + "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": "cc2512bbdc8b6ad3be036f8b69b0da92" + }, + { + "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": "eab016ae7a2fc3d794217ef409ceedd3" + }, + { + "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": "ccea116d54228efa1968e131814a2a6f" + }, + { + "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": "bec8028611f5f753ab1237558b5cf5dd" + }, + { + "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": "7d2af1d5b2fc2f0aba23d778e2297904" + }, + { + "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": "cbcb1ca06eb5395e065f45ee2433b1a7" + } + ] +} \ No newline at end of file