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": 31066067264.0, + "BitsPerParam": 3.4852341809509815 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 105012480, + "records": [ + { + "name": "lm_head.q_weight", + "shape": [ + 32016, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 105012480, + "byteOffset": 0 + } + ], + "md5sum": "3b77da8e825b28ff8d9ad1c89d704cb7" + }, + { + "dataPath": "params_shard_1.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": "072c6097e45fa88d5c6fa0f9ded04fd6" + }, + { + "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": "03980b6d8cd2649dc680c7ed807dcf70" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.78.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "4184ce5c051e09a176fab90c26b1f226" + }, + { + "dataPath": "params_shard_4.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": "04e34d666e244194422dbd42bc4ff166" + }, + { + "dataPath": "params_shard_5.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": "0c60fc6872dfa45739408e28ddb44943" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 28281760, + "records": [ + { + "name": "lm_head.q_scale", + "shape": [ + 32016, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 13126560, + "byteOffset": 0 + }, + { + "name": "model.layers.78.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 13126560 + }, + { + "name": "model.layers.78.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 13142944 + }, + { + "name": "model.layers.78.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 24890272 + }, + { + "name": "model.layers.78.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 24906656 + }, + { + "name": "model.layers.79.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 28265376 + } + ], + "md5sum": "c2da6ced16055e75ce5477dc601f572d" + }, + { + "dataPath": "params_shard_7.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": "2434b8c3a1339279c77e9a644ec338ea" + }, + { + "dataPath": "params_shard_8.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": "83ef1ca86d79680fecbbd05bc3b0ee44" + }, + { + "dataPath": "params_shard_9.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": "43a3016f18dd1d882fbc582048556c71" + }, + { + "dataPath": "params_shard_10.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": "ab228dc030eb8c418950c7eb65d462ea" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 105012480, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 32016, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 105012480, + "byteOffset": 0 + } + ], + "md5sum": "8abd56caa30dbda281fab15161a3ddec" + }, + { + "dataPath": "params_shard_12.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": "7bf243252ddd33f66ced40e51f454438" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 32480160, + "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.embed_tokens.q_scale", + "shape": [ + 32016, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 13126560, + "byteOffset": 19337216 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 32463776 + } + ], + "md5sum": "2643a635e5485c5aa23b37b46ca045f7" + }, + { + "dataPath": "params_shard_14.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": "c8640b08100fa39223ec8f9b7148a108" + }, + { + "dataPath": "params_shard_15.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": "c123913fb9fac12c58c12f9798c49c4c" + }, + { + "dataPath": "params_shard_16.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": "05136929f6491284900f74f026f64cdc" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "ca814d10e62b07744f4fb300af7dbb86" + }, + { + "dataPath": "params_shard_18.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": "daf7473b0c2e6539302524e4e93a403a" + }, + { + "dataPath": "params_shard_19.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": "525c7fc10ec6e3002f276339efac712c" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 23511040, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 0 + } + ], + "md5sum": "1357e215abd034d5b860c0752ada05bc" + }, + { + "dataPath": "params_shard_21.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": "94ce3d9f7723b05cda2b8898c26bf4ae" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "3f094bb0c44128c830663cb0915125c8" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "170554cbd4c8b61d0c6e8c953cc94724" + }, + { + "dataPath": "params_shard_24.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": "0879857cf97ad524bd5129cad4b187bd" + }, + { + "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": "47aaf6d96477ad57db3744aaf8aa843c" + }, + { + "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": "2e8da988f6a196565eadb65edf0fb3ad" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 31068160, + "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.2.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "c79df70f92fcc63b335bac807a050323" + }, + { + "dataPath": "params_shard_28.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": "ecedfa720a94d02596b000cdcb70dedb" + }, + { + "dataPath": "params_shard_29.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": "787fbdb3bd20bafc2879d148240ab393" + }, + { + "dataPath": "params_shard_30.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": "6a7e426aef2562110a8fc39992cfb2b4" + }, + { + "dataPath": "params_shard_31.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": "4f73b2740682010e0f5c0bed258bc5db" + }, + { + "dataPath": "params_shard_32.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": "acb926315e79b36fbf2f5947278e61dd" + }, + { + "dataPath": "params_shard_33.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": "e30234d0ff57e031e446fdf33df4cf27" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 22712320, + "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.10.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "922a059e2bdb06c0ef791f43f46ded34" + }, + { + "dataPath": "params_shard_35.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": "6a0b6b229d410ae15c2b21e288ac4975" + }, + { + "dataPath": "params_shard_36.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": "daeada7f13e3b890e0b3130eceb3f9c3" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a4b77f1e42577245603fcc3a126707de" + }, + { + "dataPath": "params_shard_38.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": "389287981513555f78075cce50b40c57" + }, + { + "dataPath": "params_shard_39.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": "ff637db501211b35085ea76ea70e84b1" + }, + { + "dataPath": "params_shard_40.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": "535c3e71689ee49aac31b9d7c0c54fdc" + }, + { + "dataPath": "params_shard_41.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": "41167d8126f951d6694ea1e22934f02b" + }, + { + "dataPath": "params_shard_42.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": "74ef19b2a21418357da7c50c3ce459cd" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 11763712 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15138816 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26886144 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26902528 + } + ], + "md5sum": "5205f6a44135bc1c629c90887f81bfbe" + }, + { + "dataPath": "params_shard_44.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": "2d5067b000a7f73c0612273ab4c97932" + }, + { + "dataPath": "params_shard_45.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": "4b76c958f7900ad2957900733b121c11" + }, + { + "dataPath": "params_shard_46.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": "1282e11889df2811bb76ed28e15f8e87" + }, + { + "dataPath": "params_shard_47.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": "e4d22f0335050b0e64c03ac09a78e9c2" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "217c01be7f2e58431e6b4c34295d870d" + }, + { + "dataPath": "params_shard_49.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": "f7a47a63be3b81317c696700f37e9cfb" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "8ca8d58857d6745c627fe826acaeeb9b" + }, + { + "dataPath": "params_shard_51.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": "c5ca5720b31c510e3331afda4731dd85" + }, + { + "dataPath": "params_shard_52.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": "801196bd43c6a593ac7956e5cc342902" + }, + { + "dataPath": "params_shard_53.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": "1c0f07ca6bceead68e406f30fde7223e" + }, + { + "dataPath": "params_shard_54.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": "08d7bac7eaf005256a69155f56a0dcd2" + }, + { + "dataPath": "params_shard_55.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": "2713a154d1ae43e4fc92caf80c6231ce" + }, + { + "dataPath": "params_shard_56.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": "7843d49a5088ff58d8d6c0c0a06e5dcb" + }, + { + "dataPath": "params_shard_57.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": "4813d4b2a7e53c9ba69637253cb4464c" + }, + { + "dataPath": "params_shard_58.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": "0026113a7340c044817c357358e031b1" + }, + { + "dataPath": "params_shard_59.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": "86e0f827bc7f7d685f4aee89c0a1ca0c" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 26894336, + "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.12.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "a150a597f80396475eed2990404e394b" + }, + { + "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": "5ed448577af5c70c92c73220d44f4757" + }, + { + "dataPath": "params_shard_62.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": "a5bf0dfc9e98a2029da14183938b8445" + }, + { + "dataPath": "params_shard_63.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": "ec91b498d851d335f5a0eb88b7cf48f5" + }, + { + "dataPath": "params_shard_64.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": "eeb7985ca13c1a86c61d6ed2326bec9d" + }, + { + "dataPath": "params_shard_65.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": "e3f46f6e5a6bfa9c966179ef65023118" + }, + { + "dataPath": "params_shard_66.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": "890882b833a54b31df34fc76178bc5fc" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "160309b8bc3cb94b8e1a3af52903561a" + }, + { + "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": "3e9e38d88906adea916e7f6ef5d203ef" + }, + { + "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": "2570bba9db2e147abac71a0c3f5b7136" + }, + { + "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": "abc519ac717884a816726a74c1185e5d" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "17c8fbd44de97773377962184361981b" + }, + { + "dataPath": "params_shard_72.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": "e6ec1a0052eab9c02f4b315cbbc0f8b8" + }, + { + "dataPath": "params_shard_73.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": "3602ed2a142e59bff90b226f6c9ba15a" + }, + { + "dataPath": "params_shard_74.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": "80b6d896c1b91c0e4002f7fc59b9f7bd" + }, + { + "dataPath": "params_shard_75.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": "f72b0745263b5c356883e0160ed681fb" + }, + { + "dataPath": "params_shard_76.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": "41d842249977e2bf76e8fe5ab96c6c75" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "f80fd0ae6391cc8bb4ec94f8f25a2924" + }, + { + "dataPath": "params_shard_78.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": "e4415762a0e04c269dd2f88e00ca1e27" + }, + { + "dataPath": "params_shard_79.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": "c9bfa5d52d604814058a1914856d569a" + }, + { + "dataPath": "params_shard_80.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": "dc5738c0947abf6d9e4f787c2a6cb101" + }, + { + "dataPath": "params_shard_81.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": "59c7a0feb56b6eb0a079c233e52998e2" + }, + { + "dataPath": "params_shard_82.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": "0bd907a2e05341c63c44bf21887eff95" + }, + { + "dataPath": "params_shard_83.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": "21f0d1b34be45515a710b83a5803f755" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "931a8313889f11216189b2a44c7acb22" + }, + { + "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": "4aca0528fc3c3fc9f454153aec6d6129" + }, + { + "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": "d52878d207aa4796b6e54bfc9e82cb04" + }, + { + "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": "f354077a51b838d77d7b29772c0b7b2f" + }, + { + "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": "c522543f29278fdf55e39d3d6e93b32c" + }, + { + "dataPath": "params_shard_89.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": "9aab9003206bdaa5ab9b2a8a26e29678" + }, + { + "dataPath": "params_shard_90.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": "55f1812dfb7a77153c0896b230f66e0f" + }, + { + "dataPath": "params_shard_91.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": "894368fe49fc124d3c14b1de5f5d9200" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "5c337d14eb24b26af1e359c1772b0276" + }, + { + "dataPath": "params_shard_93.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": "d86fb452dc9f1da733c6b118463b8e63" + }, + { + "dataPath": "params_shard_94.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": "b6e97175afb26511a984768361533b71" + }, + { + "dataPath": "params_shard_95.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": "f6b1482f532c6c173a136cca6d066fdb" + }, + { + "dataPath": "params_shard_96.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": "a435ac42cd2b6ce53e619e12db762b55" + }, + { + "dataPath": "params_shard_97.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": "741f505bc1945d010bbeecd412a13c94" + }, + { + "dataPath": "params_shard_98.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": "9c5cc5fa93724cc2be5a866cf4d84c07" + }, + { + "dataPath": "params_shard_99.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": "6c7bb68d0acaad4fb7a3b4433762698c" + }, + { + "dataPath": "params_shard_100.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": "0ca24e259844092d6fa53db5e84ccc51" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 31117312, + "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.20.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "9beb2c6c38143746092444774e02972f" + }, + { + "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": "2af17b47bb67e264d3878946b20f01db" + }, + { + "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": "9d39cedc915935e727342c6aa1a99058" + }, + { + "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": "b2949daf7a6edc5850b9d30069352dea" + }, + { + "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": "745fa186ecfc8006df85653301235728" + }, + { + "dataPath": "params_shard_106.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": "f46cfc4d9e084784102b31981e8c93db" + }, + { + "dataPath": "params_shard_107.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": "3dbce68b03276ce5dee3f78538d98b65" + }, + { + "dataPath": "params_shard_108.bin", + "format": "raw-shard", + "nbytes": 23535616, + "records": [ + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23519232 + } + ], + "md5sum": "19a13b75ce8bd9ffc0c59a20b46da590" + }, + { + "dataPath": "params_shard_109.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": "b0f265d8ad3ee19fd4ed230ae4705fd7" + }, + { + "dataPath": "params_shard_110.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": "7cec8eb9111574f3fd983b78590e8c21" + }, + { + "dataPath": "params_shard_111.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": "be14a224048873d2c4af4e9464d7873c" + }, + { + "dataPath": "params_shard_112.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": "87df5c27c1ce0cb0aa78bca43d6a0314" + }, + { + "dataPath": "params_shard_113.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": "509892e49d760596d7a308967392c432" + }, + { + "dataPath": "params_shard_114.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": "174547f9fae17970ffa27783b027a7f1" + }, + { + "dataPath": "params_shard_115.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "7c99432d45e45d944e27c3c65a9aeea8" + }, + { + "dataPath": "params_shard_116.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": "3bd1420382aecc48f054898dd3191763" + }, + { + "dataPath": "params_shard_117.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": "3fa42b82c95a8b4c83a5b8d6d44743aa" + }, + { + "dataPath": "params_shard_118.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": "b44c048ff9262cdeac9edd5657337286" + }, + { + "dataPath": "params_shard_119.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": "b71eb1160fc556d3849e50517d12ad9c" + }, + { + "dataPath": "params_shard_120.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": "ff588b208b42223a1ce76f8b6edb7bc3" + }, + { + "dataPath": "params_shard_121.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": "538458bf4c03ce22629fb21781587bce" + }, + { + "dataPath": "params_shard_122.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": "3b597ac7f87f44ae9825011e67a0bc69" + }, + { + "dataPath": "params_shard_123.bin", + "format": "raw-shard", + "nbytes": 26894336, + "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.4.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "bc1b405844fc368d9efe16a6b325b230" + }, + { + "dataPath": "params_shard_124.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": "8cb74d75649917d32c20e8b89d99f6db" + }, + { + "dataPath": "params_shard_125.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": "973513d1b9f8cc54e4dbf08b097ccc5d" + }, + { + "dataPath": "params_shard_126.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": "eb79cb20d1e0d277f38b8a4f51be1d7c" + }, + { + "dataPath": "params_shard_127.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": "d84c7cb36d7162365acd3b1d8b9e6a53" + }, + { + "dataPath": "params_shard_128.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": "f29f8e7f8102be0e8ac0db6f7879a7a0" + }, + { + "dataPath": "params_shard_129.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": "3c8de83ee08e10cd85644a67d1d99768" + }, + { + "dataPath": "params_shard_130.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": "562640b97d783c838ee2d801481b2930" + }, + { + "dataPath": "params_shard_131.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "75e9b9f2f00bf2deabbe30d32ced71d8" + }, + { + "dataPath": "params_shard_132.bin", + "format": "raw-shard", + "nbytes": 31100928, + "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.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 11763712 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15138816 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26886144 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26902528 + } + ], + "md5sum": "6f4334f55d081b174d319ee14f06da3a" + }, + { + "dataPath": "params_shard_133.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": "3f3ea854deb874b95cabd147370d1b79" + }, + { + "dataPath": "params_shard_134.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": "59ca1a91b5bc97fddcb52968b1976012" + }, + { + "dataPath": "params_shard_135.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": "12563f69a7107cbe9e1ea42896697864" + }, + { + "dataPath": "params_shard_136.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": "214a3b6864d27dc83f09e915f3fed4e3" + }, + { + "dataPath": "params_shard_137.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": "4bcb4ea62785d0c217c54d25fde474ff" + }, + { + "dataPath": "params_shard_138.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": "5bbd0a9ba53566945a8632373264bb08" + }, + { + "dataPath": "params_shard_139.bin", + "format": "raw-shard", + "nbytes": 22712320, + "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.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "d92e19ca2f5ac74702a36789ac37895e" + }, + { + "dataPath": "params_shard_140.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": "0cca401bd89f557eda1005f0ac0b0fb5" + }, + { + "dataPath": "params_shard_141.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": "462874e93753f36248495f29ed8b92aa" + }, + { + "dataPath": "params_shard_142.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": "c012fa1b9d95ce620f871bb5e70359b7" + }, + { + "dataPath": "params_shard_143.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": "7e67dd3ec834fc1ede854e1310426a19" + }, + { + "dataPath": "params_shard_144.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": "5f7dbb8d8b96c1ae814d35b24b009a43" + }, + { + "dataPath": "params_shard_145.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": "1413cd05cb10fffa510fc97324a007d5" + }, + { + "dataPath": "params_shard_146.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": "bfb1387d74a16ec6f8eff22111fbb09d" + }, + { + "dataPath": "params_shard_147.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": "92684fd2664c0b612f85fcbe4b93d068" + }, + { + "dataPath": "params_shard_148.bin", + "format": "raw-shard", + "nbytes": 31100928, + "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.25.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "10482d4c16fe4479babff884870284ff" + }, + { + "dataPath": "params_shard_149.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "6717d310a34db65f3c5952f8875942e7" + }, + { + "dataPath": "params_shard_150.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": "7bbe04dbff98964068c5ac1b8140f084" + }, + { + "dataPath": "params_shard_151.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": "1187413e3ad92cb576a389f67cf514d5" + }, + { + "dataPath": "params_shard_152.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": "043720c6a48beb83e6cb8cdb69db84ab" + }, + { + "dataPath": "params_shard_153.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "c44fc82b3f0298ec1fe4f9a5a3b1431c" + }, + { + "dataPath": "params_shard_154.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": "fb3dbd8dbdaf0404d2cb6c198c9f5c50" + }, + { + "dataPath": "params_shard_155.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": "e140024a6220e359bb4d1d6b4b5d1284" + }, + { + "dataPath": "params_shard_156.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": "06b445db4f9d0cbceeeb367320f2024c" + }, + { + "dataPath": "params_shard_157.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": "12596a88c00170c00218f767f96a177e" + }, + { + "dataPath": "params_shard_158.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": "8a3c5a93be14b38df13957123be84b61" + }, + { + "dataPath": "params_shard_159.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": "cb7d28fa4b0cd1767467271937f4a08a" + }, + { + "dataPath": "params_shard_160.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15138816 + }, + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15155200 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26902528 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26918912 + } + ], + "md5sum": "9d51a4f200a4be7f32b8fca1ca8669aa" + }, + { + "dataPath": "params_shard_161.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": "b8b3619a4e36d4376d505e8d0474056d" + }, + { + "dataPath": "params_shard_162.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": "8275748fb187213cee7e507a7ad80d08" + }, + { + "dataPath": "params_shard_163.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": "8cc2e133f53b72e0e4018048ce1c9b9e" + }, + { + "dataPath": "params_shard_164.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": "79cfdd42114c2d79f7796a3db15e54d1" + }, + { + "dataPath": "params_shard_165.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": "a02c3dfabc6c47fcb3e0322f36067990" + }, + { + "dataPath": "params_shard_166.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": "a806ced58f23c375a6e47feb956ff499" + }, + { + "dataPath": "params_shard_167.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": "562669ed86d94a3f3c4e0a474cad4838" + }, + { + "dataPath": "params_shard_168.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": "af190aa64fc13ae6c9dea05ab46a5672" + }, + { + "dataPath": "params_shard_169.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": "2f1a22df4423bcfe9b5445407daad0c2" + }, + { + "dataPath": "params_shard_170.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": "b2bdc0b3948afd3fe08d5fadfffe0d82" + }, + { + "dataPath": "params_shard_171.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "a46f39e5c8d6ba541b005f25ffcde35b" + }, + { + "dataPath": "params_shard_172.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": "af9c9923e477049fa31baf0fe75ff190" + }, + { + "dataPath": "params_shard_173.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": "e40316666d10c89a4d87fd619b866e4b" + }, + { + "dataPath": "params_shard_174.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": "b42f7dacbbddaf2d4b8ff85a213a3d43" + }, + { + "dataPath": "params_shard_175.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": "db2ad3c285e928fe5c8d32437ae7932c" + }, + { + "dataPath": "params_shard_176.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "c355de229703ec5fc26a893a57c110d1" + }, + { + "dataPath": "params_shard_177.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": "251dfcdcdf8176862e82d1fcf39d789f" + }, + { + "dataPath": "params_shard_178.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.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "b0bfaccc26f0238591d28cf8b9bd39e3" + }, + { + "dataPath": "params_shard_179.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": "067da360fed77028c5a1e3b3b51e889e" + }, + { + "dataPath": "params_shard_180.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": "fd6a30c23c811129d52de2cb408fc1bf" + }, + { + "dataPath": "params_shard_181.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": "152a46143f5fe4f424dc7ab50a159cb6" + }, + { + "dataPath": "params_shard_182.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": "6c4b6c0614a9ae6500e6cbb44553e499" + }, + { + "dataPath": "params_shard_183.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": "3856ac432db9a1942e5f7aa330909d83" + }, + { + "dataPath": "params_shard_184.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": "40543eee8555ea24d32d6dc9b0cf21c8" + }, + { + "dataPath": "params_shard_185.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": "aebfb87df64fd49c3870957c52714228" + }, + { + "dataPath": "params_shard_186.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "d5a1838b7e0a6a77af00dae0a4e72a2c" + }, + { + "dataPath": "params_shard_187.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": "0899f77e808585f7dcbd9dbbccb9fa21" + }, + { + "dataPath": "params_shard_188.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": "2a4688859fcba9432582bf0b7a78ac7b" + }, + { + "dataPath": "params_shard_189.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": "ddeaf82b030a31ccfa99203a99c4fb14" + }, + { + "dataPath": "params_shard_190.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": "0dfbfe9b381f271f4b4c781524da82b6" + }, + { + "dataPath": "params_shard_191.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": "b539f86acca3161997e295e42e28b517" + }, + { + "dataPath": "params_shard_192.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": "c35ec33b731c1ca47bca7ee1dcca0a0e" + }, + { + "dataPath": "params_shard_193.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": "f605524df15db5f060facdf8e5eb496d" + }, + { + "dataPath": "params_shard_194.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": "606b25fce11336dc83d68309cad137c1" + }, + { + "dataPath": "params_shard_195.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": "e8a40110a0e78761db410d1be14684b1" + }, + { + "dataPath": "params_shard_196.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": "d29a92145d95006b0497df4b6ac84c3b" + }, + { + "dataPath": "params_shard_197.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": "05b67f7d963f3ca5022c72f2d3f59638" + }, + { + "dataPath": "params_shard_198.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": "71b17d37f06842ae8d5af3771bdf8d8b" + }, + { + "dataPath": "params_shard_199.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": "2880922b50248fad3477042f98337d23" + }, + { + "dataPath": "params_shard_200.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": "d06f1a45e585e65b28d27eb4f77f3f0c" + }, + { + "dataPath": "params_shard_201.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": "1468185d80ab5d41024c90ed5294872f" + }, + { + "dataPath": "params_shard_202.bin", + "format": "raw-shard", + "nbytes": 23535616, + "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.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23519232 + } + ], + "md5sum": "4f9368aa11e0ebb170d73a7c6f070e9e" + }, + { + "dataPath": "params_shard_203.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": "168a2e8e9851bf554963e8ee08d4ba69" + }, + { + "dataPath": "params_shard_204.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": "5cd841b6f56b23b0181aa2d8635d16c5" + }, + { + "dataPath": "params_shard_205.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": "230a319a3965de7f6f1c8e1caddc1a5e" + }, + { + "dataPath": "params_shard_206.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": "c5e2c779337613182b1372fd816fd060" + }, + { + "dataPath": "params_shard_207.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": "921e2df02ec3712f96cb3f89e0358739" + }, + { + "dataPath": "params_shard_208.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": "16daff247d09f72becbb7066c2fd0f7d" + }, + { + "dataPath": "params_shard_209.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": "b6e6f85258342cbf3f81892946db78a3" + }, + { + "dataPath": "params_shard_210.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": "3d409b7af6170485717256191e314c8f" + }, + { + "dataPath": "params_shard_211.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.36.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 11763712 + }, + { + "name": "model.layers.37.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.37.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15138816 + }, + { + "name": "model.layers.37.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26886144 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26902528 + } + ], + "md5sum": "e21672d13946b4d8d70b40539c6b3a36" + }, + { + "dataPath": "params_shard_212.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": "b516aa8c2a1a4e182e8513e59573bca0" + }, + { + "dataPath": "params_shard_213.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": "4514ae95f68ec979f6218b1db0d75d45" + }, + { + "dataPath": "params_shard_214.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": "96d8d3988da97d846461c5d582bffefe" + }, + { + "dataPath": "params_shard_215.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": "421fb9561ec74c66bdde286c8ab03cba" + }, + { + "dataPath": "params_shard_216.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": "7d8419ffa351e652b42d80d8c85269d6" + }, + { + "dataPath": "params_shard_217.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": "cece84df8b038ecc8eb10abd669663aa" + }, + { + "dataPath": "params_shard_218.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.37.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.38.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.38.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.38.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.39.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "02733d890f57afaef7ec8f8dcc55eb7b" + }, + { + "dataPath": "params_shard_219.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": "8bb5dc24609da76877c926056e4c95fd" + }, + { + "dataPath": "params_shard_220.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": "1f99f8f328a8611da537d633d799da32" + }, + { + "dataPath": "params_shard_221.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": "ea617fe9daa808e38bd4d5dc931488bb" + }, + { + "dataPath": "params_shard_222.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.39.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "a8e341deaafbc17e8b2e8a75d73f9dbb" + }, + { + "dataPath": "params_shard_223.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": "05d6255c9ea85b36888b510cb5e27883" + }, + { + "dataPath": "params_shard_224.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": "b9df47871d27165108197096ee312b07" + }, + { + "dataPath": "params_shard_225.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": "67cbd1ae6aad806ddf655f7ffce60b2f" + }, + { + "dataPath": "params_shard_226.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": "ba90cbc308cc0948df86eba65f93e9b8" + }, + { + "dataPath": "params_shard_227.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.39.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.39.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.40.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.40.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.40.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "7ddd7c50ce0bd6635760e82f184e502b" + }, + { + "dataPath": "params_shard_228.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.40.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.40.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "f08d8fced57ee1b3b95260599f45b5aa" + }, + { + "dataPath": "params_shard_229.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": "2af55b832e12dc72606a8dc0d5f069a7" + }, + { + "dataPath": "params_shard_230.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": "76add1a24bb3005d24f5993da19eddc3" + }, + { + "dataPath": "params_shard_231.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": "3781ca58f59918e8910faa28ce3bee2a" + }, + { + "dataPath": "params_shard_232.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.40.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.41.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "00c01a49d46099e20b023d54906029ec" + }, + { + "dataPath": "params_shard_233.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": "64a8296e1da4b4bd8162ccb38d1b01cf" + }, + { + "dataPath": "params_shard_234.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": "8b0a11f49593a651db79e3d24fea092f" + }, + { + "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": "9775f93e1ef3301c6960c72b9d5f04ac" + }, + { + "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": "11c1a7b5c6776293525bbe7aec61566d" + }, + { + "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": "a2118c51e0bc7ad8fc16e3b98123255c" + }, + { + "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": "ecb0468dd264cbbae8fd673d32d7b237" + }, + { + "dataPath": "params_shard_239.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.41.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.41.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.41.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.41.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.42.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15138816 + }, + { + "name": "model.layers.42.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15155200 + }, + { + "name": "model.layers.42.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26902528 + }, + { + "name": "model.layers.42.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26918912 + } + ], + "md5sum": "f8b84db84992a5dcbfa64fc384056d53" + }, + { + "dataPath": "params_shard_240.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": "f4bfbbc9357be40fa8ca1d56b1bed7af" + }, + { + "dataPath": "params_shard_241.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": "11d915422b5abb5473d89061c346be3a" + }, + { + "dataPath": "params_shard_242.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": "5494699b72d12c193279d59ec014ea57" + }, + { + "dataPath": "params_shard_243.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": "e152b6e4d382a042420d0777a20541eb" + }, + { + "dataPath": "params_shard_244.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.43.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "49d33739d10c7439b7fe1ee54e72a624" + }, + { + "dataPath": "params_shard_245.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": "f8b6a361118c80e96ab2cae9d073419f" + }, + { + "dataPath": "params_shard_246.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": "4648a26419f4fe02cabad58a9f83329c" + }, + { + "dataPath": "params_shard_247.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": "8370afbad96840a9ee69a6b97fdfdb84" + }, + { + "dataPath": "params_shard_248.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": "4cf6cc13d3f44d01fec177d91c1d3dbd" + }, + { + "dataPath": "params_shard_249.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": "8188a66438a956c48b836a342580d623" + }, + { + "dataPath": "params_shard_250.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.42.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.43.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.43.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.43.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.43.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.43.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.44.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.44.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.44.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "89a88a54cb3049281246cda5fc69b92f" + }, + { + "dataPath": "params_shard_251.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": "6343e8bef47947c96abc0b448b531c97" + }, + { + "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": "1a0c08f9992433cbe2dfc4970a246770" + }, + { + "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": "69b272d8a0839210e0721bd87d565dd3" + }, + { + "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": "644e3903e491b6ecd88c19c1df61832b" + }, + { + "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": "7715b842809fe1e7d0da3af93078af83" + }, + { + "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": "0ab75c4062b9ab1eac398f4999e0c0ce" + }, + { + "dataPath": "params_shard_257.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.44.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.44.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.45.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.45.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.45.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.45.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.45.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.46.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "57775206533dcfd0bb351c02e06a5fe0" + }, + { + "dataPath": "params_shard_258.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": "42b4e66c05bdc763ce258e36ce63c771" + }, + { + "dataPath": "params_shard_259.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": "8e96e832f833ab7ac5623af1c5af4c54" + }, + { + "dataPath": "params_shard_260.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": "46f601fb9a44587adbb4d509d02c3059" + }, + { + "dataPath": "params_shard_261.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.46.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "db3badce110b942c5413fde8b630eb7d" + }, + { + "dataPath": "params_shard_262.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": "d0f16bd9fa62fd5d769e9834afed4772" + }, + { + "dataPath": "params_shard_263.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": "b574296863819961254f3f9a1a994f34" + }, + { + "dataPath": "params_shard_264.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": "b11c9fea4aef3195c4e23ac32011bcf6" + }, + { + "dataPath": "params_shard_265.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.46.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.46.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.46.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.46.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.47.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.47.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.47.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "abf514336db46605dfbb28840fac6ed8" + }, + { + "dataPath": "params_shard_266.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": "2989bb7d2c4f3d7189df4767a02dbe83" + }, + { + "dataPath": "params_shard_267.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": "2c8b545d272a3f6e2ff0996edfdec0f8" + }, + { + "dataPath": "params_shard_268.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": "3ec020c946178fd0c86f103e82c580bd" + }, + { + "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": "94679dce159c7bb1d3d3c3d0a060f27b" + }, + { + "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": "a9ed2688228344df26a441730e8e3fc1" + }, + { + "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": "7111b5dd1ff4594a852fc3c0de92f798" + }, + { + "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": "d4874eccd66a70e9413cbecba45a8f8f" + }, + { + "dataPath": "params_shard_273.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": "398afd61dd769c30657aab56476485ed" + }, + { + "dataPath": "params_shard_274.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.47.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.47.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.48.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.48.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.48.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.48.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.48.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.49.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "734e186858c211d271f600e60778afd1" + }, + { + "dataPath": "params_shard_275.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": "9fc27c505e7fa12fa4a99058acae2061" + }, + { + "dataPath": "params_shard_276.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": "668405fcb7d25b1dbcb22c3eac723132" + }, + { + "dataPath": "params_shard_277.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": "1527c9a3a77d89b78362c61b1dc7ce5a" + }, + { + "dataPath": "params_shard_278.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": "9d0508448c8268a5da16716e261fc32c" + }, + { + "dataPath": "params_shard_279.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": "738d018e6212c665fcfcec20c97327fc" + }, + { + "dataPath": "params_shard_280.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": "95289d377cdf3bfc1e6bfdc70de49771" + }, + { + "dataPath": "params_shard_281.bin", + "format": "raw-shard", + "nbytes": 23535616, + "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.49.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.49.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.50.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23519232 + } + ], + "md5sum": "b1d9b120ec7a40fc6ba3964e40fc8d3d" + }, + { + "dataPath": "params_shard_282.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": "69d237a75e2f9657dc561034d3f4c1ba" + }, + { + "dataPath": "params_shard_283.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": "2d13e5b5b9523db266879934d7d9127a" + }, + { + "dataPath": "params_shard_284.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": "63a2dad390d71d3752f4dc06ffb3f334" + }, + { + "dataPath": "params_shard_285.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": "eec9490e5d6921d85b51d619a9b839d7" + }, + { + "dataPath": "params_shard_286.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": "88d62eacaad1965f8615e06e3577be68" + }, + { + "dataPath": "params_shard_287.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": "22f50323894113fda50991fb089302aa" + }, + { + "dataPath": "params_shard_288.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": "3a06a48fd7360b5ca18362c93ce45924" + }, + { + "dataPath": "params_shard_289.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": "73667068dad461d3727ab5c8bbc54cb3" + }, + { + "dataPath": "params_shard_290.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "f84f64c87d247941a2f647360b1d7270" + }, + { + "dataPath": "params_shard_291.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": "b376549d6b1b6dd05fd3410e6ec2c884" + }, + { + "dataPath": "params_shard_292.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": "e87f7951430281c662840a7da28f89a3" + }, + { + "dataPath": "params_shard_293.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": "acd61748d11f5823a1937637cc53c528" + }, + { + "dataPath": "params_shard_294.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": "b36398c418dbc6267568284bd08f7500" + }, + { + "dataPath": "params_shard_295.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": "1a6607580fc2731f4c0ce61844e826e6" + }, + { + "dataPath": "params_shard_296.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": "cc3d755c87a23493e85c4abe12be4008" + }, + { + "dataPath": "params_shard_297.bin", + "format": "raw-shard", + "nbytes": 23535616, + "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.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.50.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23519232 + } + ], + "md5sum": "769b3eacc36b488c240a1757f4e3ad45" + }, + { + "dataPath": "params_shard_298.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": "e285ebbfcfe654401e50c9750780dfdc" + }, + { + "dataPath": "params_shard_299.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": "afc8ca194acee1439379e325a5d0bab0" + }, + { + "dataPath": "params_shard_300.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": "92a308b1ef3ba942a8044abea7737e4d" + }, + { + "dataPath": "params_shard_301.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": "358101eab1f75e1dbfa99e942b885634" + }, + { + "dataPath": "params_shard_302.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": "e98bbdd76bffef02ba46901f16a4260b" + }, + { + "dataPath": "params_shard_303.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": "3cd3c7b1052e17e32884ecde77191636" + }, + { + "dataPath": "params_shard_304.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": "d4146c7bc761a6b076663c68cff18380" + }, + { + "dataPath": "params_shard_305.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": "4f4303991167af9c8d1312cb05602828" + }, + { + "dataPath": "params_shard_306.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.50.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.50.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.50.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 11763712 + }, + { + "name": "model.layers.51.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.51.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15138816 + }, + { + "name": "model.layers.51.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26886144 + }, + { + "name": "model.layers.51.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26902528 + } + ], + "md5sum": "c3ad8d2e4b55a47bb5a0237c1d02279e" + }, + { + "dataPath": "params_shard_307.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": "d1816dd485e6dffae20a73c77b995139" + }, + { + "dataPath": "params_shard_308.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": "97848348678a9a277be4de5746386474" + }, + { + "dataPath": "params_shard_309.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": "48501593a1809f5f4702e4b284543424" + }, + { + "dataPath": "params_shard_310.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": "1f43a89089c65f97850f77e83a898857" + }, + { + "dataPath": "params_shard_311.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": "a713c7f47b36f9aecd1bce061dd082a1" + }, + { + "dataPath": "params_shard_312.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": "fabe9ac848988dbe27484d52acb6c0c3" + }, + { + "dataPath": "params_shard_313.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.51.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.52.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.52.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.52.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.52.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.52.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.53.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "c08c0574da457cc7fae4dbdd6b04f66b" + }, + { + "dataPath": "params_shard_314.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": "167b45540d1307c6662630be3a8dddaa" + }, + { + "dataPath": "params_shard_315.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": "88696dfc1d7ae6fba15f2e0f42cde0c6" + }, + { + "dataPath": "params_shard_316.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": "97516801b739e23320677c86f18da185" + }, + { + "dataPath": "params_shard_317.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": "eecf7c3e665d05b283916ee62d88678a" + }, + { + "dataPath": "params_shard_318.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": "6a9627210e27760cfe7f3c7ff9d807e6" + }, + { + "dataPath": "params_shard_319.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": "ae8dd61aae4b012e1e1347acc2ef6211" + }, + { + "dataPath": "params_shard_320.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": "ee98f81bedf1b99a3bf3cbcc3e9b9d9b" + }, + { + "dataPath": "params_shard_321.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": "3cda934ca1f8d802c95601fdf3d93345" + }, + { + "dataPath": "params_shard_322.bin", + "format": "raw-shard", + "nbytes": 31100928, + "records": [ + { + "name": "model.layers.53.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.53.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.53.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.53.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.54.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.54.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.54.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "98ebf8d57ac857f7428687ae59784685" + }, + { + "dataPath": "params_shard_323.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.54.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.54.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "8fe937806b3c3565ed68a7d2b9eeef7e" + }, + { + "dataPath": "params_shard_324.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": "29eb85ba9f7f7f0732e8085746006a4d" + }, + { + "dataPath": "params_shard_325.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": "e942f2aa23776b66594e0fe1a0873766" + }, + { + "dataPath": "params_shard_326.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.55.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "5497190ee9cb61e761b845a968f7aca4" + }, + { + "dataPath": "params_shard_327.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.54.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.55.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.55.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "2d3ad54d0ffde2a6e09003d61aea5eca" + }, + { + "dataPath": "params_shard_328.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": "c0a29792eb5f11b759730e3d43890475" + }, + { + "dataPath": "params_shard_329.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": "b458eeacf9e3891abf57d8d269376f2a" + }, + { + "dataPath": "params_shard_330.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": "9587c44d4919173361ccbe6986d24236" + }, + { + "dataPath": "params_shard_331.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": "bf15165fd19822945f0aefde4a80eb80" + }, + { + "dataPath": "params_shard_332.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": "4d0909ed29e0d86e1be66fe496971ea3" + }, + { + "dataPath": "params_shard_333.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": "d07973561c43331a35901edc3e86fc94" + }, + { + "dataPath": "params_shard_334.bin", + "format": "raw-shard", + "nbytes": 31117312, + "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.55.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.55.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.55.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.56.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15138816 + }, + { + "name": "model.layers.56.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15155200 + }, + { + "name": "model.layers.56.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26902528 + }, + { + "name": "model.layers.56.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26918912 + } + ], + "md5sum": "8adb61fea4b4f5466b8389939657460c" + }, + { + "dataPath": "params_shard_335.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": "ad2a23228510ca7f088cc343b9be414f" + }, + { + "dataPath": "params_shard_336.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": "949b1b89ba2e0a488cd59fbc5b69d3a1" + }, + { + "dataPath": "params_shard_337.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": "2973b68c70727f23a0be7b9cc75d454b" + }, + { + "dataPath": "params_shard_338.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": "e7456b9196326c8d9867924866c99372" + }, + { + "dataPath": "params_shard_339.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": "d5abf1e157475c7475a7d27e46a580a5" + }, + { + "dataPath": "params_shard_340.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": "d2feb05ac36215c0d13873ff3b0ec4d4" + }, + { + "dataPath": "params_shard_341.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": "01e447ffd7656a5824b9ab1483d958af" + }, + { + "dataPath": "params_shard_342.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": "59dae452c1d55d1ec1f2025a61a2b392" + }, + { + "dataPath": "params_shard_343.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": "80c89ac66dd91391227420de76802f33" + }, + { + "dataPath": "params_shard_344.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": "ba4576b1e4577e4a9cad658586a80417" + }, + { + "dataPath": "params_shard_345.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.56.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.57.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.57.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.57.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.57.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.57.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.58.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.58.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.58.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "5c3df11afbdab21e1593cd7cbaa85251" + }, + { + "dataPath": "params_shard_346.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": "a8f43d96f1bb6d05f65205db8dbcdc9f" + }, + { + "dataPath": "params_shard_347.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": "db428ebd69397054ead9f89def51c6c2" + }, + { + "dataPath": "params_shard_348.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": "907e45e365e473876ebd2b86477ab3f7" + }, + { + "dataPath": "params_shard_349.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": "b2bfb5432b885724749d9187565755be" + }, + { + "dataPath": "params_shard_350.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": "9f77b9982dc9753c6710703220bb212d" + }, + { + "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": "387814bfd6dc77e0c4cbb7b3d7eead53" + }, + { + "dataPath": "params_shard_352.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.58.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.58.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.59.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.59.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.59.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.59.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.59.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.60.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "0dfa8820ddcff8e7c68f6c88a59650a2" + }, + { + "dataPath": "params_shard_353.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": "8e890a6d97bb9f9c7b88c83f844199b8" + }, + { + "dataPath": "params_shard_354.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": "6c9c07fc641f328b988b6c17cab8b731" + }, + { + "dataPath": "params_shard_355.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": "a80de85f96b092635aa045e1f1b24418" + }, + { + "dataPath": "params_shard_356.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": "f353751942f3d29b3fd22391292b1022" + }, + { + "dataPath": "params_shard_357.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": "76671958f9e8b54012617c026f299715" + }, + { + "dataPath": "params_shard_358.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": "98605ea7422eff6c0e92224d2be77938" + }, + { + "dataPath": "params_shard_359.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": "c8466237965405cf8b42fcc724b7a6f4" + }, + { + "dataPath": "params_shard_360.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.60.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.60.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.60.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.60.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.61.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.61.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.61.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "b493dfaffb5fa49b45c107bb2e8c4fb8" + }, + { + "dataPath": "params_shard_361.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": "e6ae8f3712c652f18052a38591dbb623" + }, + { + "dataPath": "params_shard_362.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": "050ddd71e6209dde9966d3b0c22826f3" + }, + { + "dataPath": "params_shard_363.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": "fc24de14a253136e236b408903419df2" + }, + { + "dataPath": "params_shard_364.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": "15126a4ee3eded8ac2e5e582f777237c" + }, + { + "dataPath": "params_shard_365.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": "b2d7d011ef1449b10817a6475b96cfe9" + }, + { + "dataPath": "params_shard_366.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": "0c26a32faa6dd487e31bdcb9ac79d91f" + }, + { + "dataPath": "params_shard_367.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.62.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "ea6fcbf8cf44aa56d49e04cbd01762c7" + }, + { + "dataPath": "params_shard_368.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": "a5938e4486284f609def030db3bc3e6e" + }, + { + "dataPath": "params_shard_369.bin", + "format": "raw-shard", + "nbytes": 31117312, + "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.62.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.62.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.62.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.62.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.62.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.63.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "1b6ee1669ed9bd280e58afdc06873ef9" + }, + { + "dataPath": "params_shard_370.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": "85633f4761eae8e8811b594f51c08b7a" + }, + { + "dataPath": "params_shard_371.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": "d6710df1b92ad83522fa3582049656bb" + }, + { + "dataPath": "params_shard_372.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": "d9c514ace21fd88ca8675b30e9f32498" + }, + { + "dataPath": "params_shard_373.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": "b079b46916580d4126217894a50ca7c3" + }, + { + "dataPath": "params_shard_374.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": "0052e8b3acfc4319d5e7e221e1ef0b6f" + }, + { + "dataPath": "params_shard_375.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": "8e238f6f261b71d98a30c39dcb7af058" + }, + { + "dataPath": "params_shard_376.bin", + "format": "raw-shard", + "nbytes": 23535616, + "records": [ + { + "name": "model.layers.63.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.63.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.63.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.63.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.64.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.64.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23519232 + } + ], + "md5sum": "87d50526eaa142020fdd891e75862c67" + }, + { + "dataPath": "params_shard_377.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": "a8f2603c44f5bba8f08bfec8ec07cac4" + }, + { + "dataPath": "params_shard_378.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": "04dc727fd14310d4c9437b109314eb38" + }, + { + "dataPath": "params_shard_379.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": "12557b08b47aefff97c2507e4b749ca5" + }, + { + "dataPath": "params_shard_380.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": "67b4159d9eeb1b2b410f6b8cbefee929" + }, + { + "dataPath": "params_shard_381.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": "9dfbe7840edb8dea4ce77748ff6fe2f7" + }, + { + "dataPath": "params_shard_382.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": "82704550e5dc71c72d6d3a25c901e4b4" + }, + { + "dataPath": "params_shard_383.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": "41b304cce24359acdc39ea6b0cbd8d0a" + }, + { + "dataPath": "params_shard_384.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": "5ed0ae7867c5384f97541e81c3db93e2" + }, + { + "dataPath": "params_shard_385.bin", + "format": "raw-shard", + "nbytes": 31100928, + "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.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 11763712 + }, + { + "name": "model.layers.65.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.65.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15138816 + }, + { + "name": "model.layers.65.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26886144 + }, + { + "name": "model.layers.65.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26902528 + } + ], + "md5sum": "2998f8d6284d63add68c770d013cb10c" + }, + { + "dataPath": "params_shard_386.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": "5c121fc53e6b1fd80cfc1ff52cf388ac" + }, + { + "dataPath": "params_shard_387.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": "dbf052ca9e7a7556618ad85e2ca83469" + }, + { + "dataPath": "params_shard_388.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": "154c58df8f8b1d5aec0aac82f3255d16" + }, + { + "dataPath": "params_shard_389.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": "8c5bae7f8741be4ccf0ecb3b87e4e714" + }, + { + "dataPath": "params_shard_390.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": "faa5c2956eb4628d8429deded05b7c34" + }, + { + "dataPath": "params_shard_391.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": "64fc11a70a79ddfe1413b08ba5d6574d" + }, + { + "dataPath": "params_shard_392.bin", + "format": "raw-shard", + "nbytes": 22712320, + "records": [ + { + "name": "model.layers.65.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.66.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.66.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.66.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.66.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.66.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.67.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 22695936 + } + ], + "md5sum": "07ba1a8351a2c8df8d8f8d349d26b68a" + }, + { + "dataPath": "params_shard_393.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": "be8c28cc252709970d508f56ac25e23e" + }, + { + "dataPath": "params_shard_394.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": "1507749a3543fb7ff36e34d393f6ec77" + }, + { + "dataPath": "params_shard_395.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": "74374fd9d68eb89a51ea28c1ff18bdd3" + }, + { + "dataPath": "params_shard_396.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": "b89b8692d0913640b9aa51b1cef6181e" + }, + { + "dataPath": "params_shard_397.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": "c8143c59b6bf91720ae64c397b70bdba" + }, + { + "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": "e758c506ff3ae11d41344d8e5a0ef92b" + }, + { + "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": "5b125082f66fd3b36d71354faba843e5" + }, + { + "dataPath": "params_shard_400.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": "1b57776625ece812fc35a4f7e69a57df" + }, + { + "dataPath": "params_shard_401.bin", + "format": "raw-shard", + "nbytes": 31100928, + "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.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 19320832 + }, + { + "name": "model.layers.68.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 19337216 + }, + { + "name": "model.layers.68.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31084544 + } + ], + "md5sum": "a7143d98f88748d530861d514e486ade" + }, + { + "dataPath": "params_shard_402.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.68.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 0 + }, + { + "name": "model.layers.68.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 4198400 + } + ], + "md5sum": "35b9eb5a67cf983ca36cf4cd9e85a050" + }, + { + "dataPath": "params_shard_403.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": "5cef94fbcb168301d25f9df4d204d763" + }, + { + "dataPath": "params_shard_404.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": "feef71be188031643cbc8d811963d40d" + }, + { + "dataPath": "params_shard_405.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": "a7da21e6bd15c25e3c77092c36a63479" + }, + { + "dataPath": "params_shard_406.bin", + "format": "raw-shard", + "nbytes": 31068160, + "records": [ + { + "name": "model.layers.68.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.69.mlp.gate_up_proj.q_scale", + "shape": [ + 57344, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 23511040, + "byteOffset": 3358720 + }, + { + "name": "model.layers.69.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26869760 + } + ], + "md5sum": "43f15ec34b83772596547e3d612d0433" + }, + { + "dataPath": "params_shard_407.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": "d4052328f4ab36d5ecf1d44ab4220547" + }, + { + "dataPath": "params_shard_408.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": "12ceacc533048876a24d30409b265c1b" + }, + { + "dataPath": "params_shard_409.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": "7a2c8bfca1f64d5d001d257e35284f9d" + }, + { + "dataPath": "params_shard_410.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": "4593aec420e33870d5a000755c59705d" + }, + { + "dataPath": "params_shard_411.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": "400271e2c16e0694814d68ff82a40321" + }, + { + "dataPath": "params_shard_412.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": "d490eb55c73eef6ca5106872273192aa" + }, + { + "dataPath": "params_shard_413.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.69.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.69.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.69.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.69.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.70.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15138816 + }, + { + "name": "model.layers.70.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 15155200 + }, + { + "name": "model.layers.70.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26902528 + }, + { + "name": "model.layers.70.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 26918912 + } + ], + "md5sum": "dd666c86254e4ac7eff07db1e5f94f60" + }, + { + "dataPath": "params_shard_414.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": "b5d7256957bf4aec0dd2bb376b86dca0" + }, + { + "dataPath": "params_shard_415.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": "87d9a58fcbc9bc1aab4b1f2d17d6e83e" + }, + { + "dataPath": "params_shard_416.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": "54ba3d3b38f6068f476be7f25fa1b1c8" + }, + { + "dataPath": "params_shard_417.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": "9132f470a721b3273b285b711db024ba" + }, + { + "dataPath": "params_shard_418.bin", + "format": "raw-shard", + "nbytes": 26869760, + "records": [ + { + "name": "model.layers.71.self_attn.o_proj.q_weight", + "shape": [ + 8192, + 820 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 26869760, + "byteOffset": 0 + } + ], + "md5sum": "497b91fc40e0440f7690fb34a94b3fd5" + }, + { + "dataPath": "params_shard_419.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": "4da988b601fe438a2ae3328417d57076" + }, + { + "dataPath": "params_shard_420.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": "84bdd79e8d43efc5df1600e537d4633a" + }, + { + "dataPath": "params_shard_421.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": "48b5e8459c455f3767ff045dcae96d5a" + }, + { + "dataPath": "params_shard_422.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": "49e3749179c97b4fdfd8caffa8134436" + }, + { + "dataPath": "params_shard_423.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": "bf739cc3df083c1f70411011ee5b55b6" + }, + { + "dataPath": "params_shard_424.bin", + "format": "raw-shard", + "nbytes": 30269440, + "records": [ + { + "name": "model.layers.70.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 0 + }, + { + "name": "model.layers.71.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 3358720 + }, + { + "name": "model.layers.71.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 3375104 + }, + { + "name": "model.layers.71.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 15122432 + }, + { + "name": "model.layers.71.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 15138816 + }, + { + "name": "model.layers.71.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 19337216 + }, + { + "name": "model.layers.72.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 22695936 + }, + { + "name": "model.layers.72.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 26894336 + }, + { + "name": "model.layers.72.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 30253056 + } + ], + "md5sum": "e0348a849380b64ebc5c5a5b34167dde" + }, + { + "dataPath": "params_shard_425.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": "4c144423fae979143d6a36162e301c84" + }, + { + "dataPath": "params_shard_426.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": "f12cc346b3c8492628da2c9cc7e6ceab" + }, + { + "dataPath": "params_shard_427.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": "8b683fdcbccd3ed36513748cfd565681" + }, + { + "dataPath": "params_shard_428.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": "afe50eafeabed7c99d8ddbda3a569c69" + }, + { + "dataPath": "params_shard_429.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": "dd6d6e1f328eeef9983067115012937d" + }, + { + "dataPath": "params_shard_430.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": "5082f9f9670e0157085201c26ddb15a3" + }, + { + "dataPath": "params_shard_431.bin", + "format": "raw-shard", + "nbytes": 31117312, + "records": [ + { + "name": "model.layers.72.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.72.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.73.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.73.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.73.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.73.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.73.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.74.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "3b56fffcb40c740bb6438de6fd481150" + }, + { + "dataPath": "params_shard_432.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": "a4c92f81626c0a045cb7464152fe2ab5" + }, + { + "dataPath": "params_shard_433.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": "7d72bbe5b5975613cf54b75a1ec68ed7" + }, + { + "dataPath": "params_shard_434.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": "1ea52905aab62c63511e1288daf97a39" + }, + { + "dataPath": "params_shard_435.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": "3c4ea90a5398ac9da8cc7c145df83d84" + }, + { + "dataPath": "params_shard_436.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": "4c7455417bdfaa149ce1ff6146b190d6" + }, + { + "dataPath": "params_shard_437.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": "47cd55d28c7bce2c535853049777a8fb" + }, + { + "dataPath": "params_shard_438.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": "a393a3fb1d99f50e07d87ffb9d8f1faf" + }, + { + "dataPath": "params_shard_439.bin", + "format": "raw-shard", + "nbytes": 26894336, + "records": [ + { + "name": "model.layers.74.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.74.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.74.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.74.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.75.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + }, + { + "name": "model.layers.75.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 23519232 + }, + { + "name": "model.layers.75.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 26877952 + } + ], + "md5sum": "f5ef1260ab221fb4bfc0f12316d2678f" + }, + { + "dataPath": "params_shard_440.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": "ef4e95d0145c07c56e04e323919d6c22" + }, + { + "dataPath": "params_shard_441.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": "8001e72ef958b1349b9fac73cafd4cdf" + }, + { + "dataPath": "params_shard_442.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": "fdadd31671834a14c20ad30582381bc2" + }, + { + "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": "960a2e34a54bdf5df9215a2e16125e19" + }, + { + "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": "ab5a3147c527a88ac34251a618d3035c" + }, + { + "dataPath": "params_shard_445.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": "7b088caca666374d2aa5001b10f9fcfd" + }, + { + "dataPath": "params_shard_446.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": "3d86440a6502b698b3414ca5f997eccc" + }, + { + "dataPath": "params_shard_447.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": "a3acd445049e7fdbd69558df44790d75" + }, + { + "dataPath": "params_shard_448.bin", + "format": "raw-shard", + "nbytes": 31117312, + "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.76.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11763712 + }, + { + "name": "model.layers.76.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 11780096 + }, + { + "name": "model.layers.76.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 23527424 + }, + { + "name": "model.layers.76.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 23543808 + }, + { + "name": "model.layers.76.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 27742208 + }, + { + "name": "model.layers.77.input_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 31100928 + } + ], + "md5sum": "f1a715d590ae17ca3ad72683cdf0c9f1" + }, + { + "dataPath": "params_shard_449.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": "55ea21dff6c0a0e3e5fd3f1f86116ac0" + }, + { + "dataPath": "params_shard_450.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": "01e1e5099e5574c6a4cd5ec5537bdbda" + }, + { + "dataPath": "params_shard_451.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": "a040c9454f9306e7fb99e97480eb9436" + }, + { + "dataPath": "params_shard_452.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": "d7c6bec9b6066c6a022165db02a75524" + }, + { + "dataPath": "params_shard_453.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": "9bfaa55f7cad67d6995283e3e55011fd" + }, + { + "dataPath": "params_shard_454.bin", + "format": "raw-shard", + "nbytes": 23519232, + "records": [ + { + "name": "model.layers.77.mlp.down_proj.q_scale", + "shape": [ + 8192, + 717 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 11747328, + "byteOffset": 0 + }, + { + "name": "model.layers.77.post_attention_layernorm.weight", + "shape": [ + 8192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 16384, + "byteOffset": 11747328 + }, + { + "name": "model.layers.77.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 11763712 + }, + { + "name": "model.layers.77.self_attn.o_proj.q_scale", + "shape": [ + 8192, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 3358720, + "byteOffset": 15962112 + }, + { + "name": "model.layers.78.self_attn.qkv_proj.q_scale", + "shape": [ + 10240, + 205 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4198400, + "byteOffset": 19320832 + } + ], + "md5sum": "13314aa8d2407e139dc046ca1841b296" + } + ] +} \ No newline at end of file