diff --git "a/tensor-cache.json" "b/tensor-cache.json" new file mode 100644--- /dev/null +++ "b/tensor-cache.json" @@ -0,0 +1,3865 @@ +{ + "metadata": { + "ParamSize": 339, + "ParamBytes": 968001536.0, + "BitsPerParam": 4.500828151065705 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 155582464, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 151936, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 155582464, + "byteOffset": 0 + } + ], + "md5sum": "4e4ec9b621dc1492b262d3614f56d118" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 26529792, + "records": [ + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 151936, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 19447808, + "byteOffset": 0 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 19447808 + }, + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 19451904 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 25743360 + } + ], + "md5sum": "b25432b895e211b606d5462a2fee80ca" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.0.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.0.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.0.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.0.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "7f147b67ea1b2dfe5dc58d10df7a5846" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.1.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.1.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.1.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.1.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "4f6158a6f42b6838f2a7f84e7343138d" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.10.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.10.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.10.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.10.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "d7f50b7c90485628d88e4b81a060f65c" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.11.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.11.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.11.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.11.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "4504f71b0c300101424066abdcd8c12d" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.12.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.12.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.12.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.12.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "84c2ebff7133f8f641b2465e0a818ed9" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.13.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.13.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.13.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.13.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "5ddbe21093c8d51d992df21d80bb2a3e" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.14.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.14.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.14.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.14.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "ed83467a72a6729a82ab580bf3671c23" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.15.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.15.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.15.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.15.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "fe949ed7b73cf5e9667ad3c1b9e8e421" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.16.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.16.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.16.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.16.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "d695302a75d9969361f8563385269e46" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.17.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.17.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.17.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.17.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "97a9f62a016943c01c867f9c9381fb19" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.18.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.18.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.18.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.18.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "914c01e9ed2b34f26ca78b4d8880827e" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.19.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.19.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.19.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.19.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "e32786733549a299d963b859a0e01d26" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.2.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.2.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.2.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.2.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "ecebb8188356a796d54a0ba702588ca5" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.20.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.20.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.20.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.20.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "3db8e5300e13b408eb7de52bf8ed0ff5" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.21.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.21.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.21.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.21.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "10e10638f29042430d26ded29091842f" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.22.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.22.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.22.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.22.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "69b3ec2354974255621cd51fc6dfd814" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.23.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.23.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.23.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.23.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "25e579acf76a76b3857613aa36f06a45" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.24.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.24.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.24.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.24.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "f43e8d05370430b4592400a4795903c5" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.25.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.25.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.25.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.25.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "f6a2c629e34530e27ee20a81030c35ab" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.26.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.26.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.26.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.26.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "cb768d2ece0954d3a933b008de88c9e5" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.27.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.27.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.27.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.27.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "0a8be09c0859d736f3f40ae278031a25" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.3.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.3.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.3.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.3.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "6cb90725786bc42097f363266addbb2b" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.4.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.4.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "8cb47390eaa7fa50201d5b9763bf30fe" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.5.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.5.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "37deefd30f4ba3acd25f84459929a6a3" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.6.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.6.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "b4e78702c99b5d4ee5732b26d21bf1dc" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.7.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.7.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "5b115ab3b3f8a56d96356b6bacf16acf" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 28320256, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.8.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.8.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + }, + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 2048, + 768 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6291456, + "byteOffset": 21242368 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 2048, + 192 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 786432, + "byteOffset": 27533824 + } + ], + "md5sum": "233ec74919f5087d3ecc1a002051b2fe" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 21242368, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 12288, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 12582912, + "byteOffset": 0 + }, + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 12288, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1572864, + "byteOffset": 12582912 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 14155776 + }, + { + "name": "model.layers.9.self_attn.k_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 14159872 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_weight", + "shape": [ + 4096, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 4194304, + "byteOffset": 14160128 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_scale", + "shape": [ + 4096, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 524288, + "byteOffset": 18354432 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18878720 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20975872 + }, + { + "name": "model.layers.9.self_attn.q_norm.weight", + "shape": [ + 128 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 256, + "byteOffset": 21238016 + }, + { + "name": "model.norm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 21238272 + } + ], + "md5sum": "24766fa894c19f583c667dff41bd8212" + } + ] +} \ No newline at end of file