| { |
| "metadata": { |
| "ParamSize": 163, |
| "ParamBytes": 695242752.0, |
| "BitsPerParam": 4.500628909972241 |
| }, |
| "records": [ |
| { |
| "dataPath": "params_shard_0.bin", |
| "format": "raw-shard", |
| "nbytes": 131334144, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_weight", |
| "shape": [ |
| 128256, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 131334144, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "c6e04a20b8de307161e2c6eebdbaf4a1" |
| }, |
| { |
| "dataPath": "params_shard_1.bin", |
| "format": "raw-shard", |
| "nbytes": 16777216, |
| "records": [ |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "26f7aad0d36183af8f72d16ce19926f1" |
| }, |
| { |
| "dataPath": "params_shard_2.bin", |
| "format": "raw-shard", |
| "nbytes": 31498240, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_scale", |
| "shape": [ |
| 128256, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 16416768, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.0.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 16416768 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 16420864 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 24809472 |
| }, |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 25858048 |
| }, |
| { |
| "name": "model.layers.0.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 27955200 |
| }, |
| { |
| "name": "model.layers.0.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 27959296 |
| }, |
| { |
| "name": "model.layers.0.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 31105024 |
| } |
| ], |
| "md5sum": "8148b9f3269cca51846dc88248db69e3" |
| }, |
| { |
| "dataPath": "params_shard_3.bin", |
| "format": "raw-shard", |
| "nbytes": 30679040, |
| "records": [ |
| { |
| "name": "model.layers.0.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.0.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.1.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2359296 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 2363392 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10752000 |
| }, |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 11800576 |
| }, |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 28577792 |
| }, |
| { |
| "name": "model.layers.1.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 30674944 |
| } |
| ], |
| "md5sum": "84efefc178dc2d1132d00d2076de53fa" |
| }, |
| { |
| "dataPath": "params_shard_4.bin", |
| "format": "raw-shard", |
| "nbytes": 32116736, |
| "records": [ |
| { |
| "name": "model.layers.1.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.1.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.1.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.1.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 5636096 |
| }, |
| { |
| "name": "model.layers.2.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 5898240 |
| }, |
| { |
| "name": "model.layers.2.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 5902336 |
| }, |
| { |
| "name": "model.layers.2.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 14290944 |
| }, |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 15339520 |
| } |
| ], |
| "md5sum": "fedaa94284edfda2d9119601d900326b" |
| }, |
| { |
| "dataPath": "params_shard_5.bin", |
| "format": "raw-shard", |
| "nbytes": 16777216, |
| "records": [ |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "7e562041b1293a109458745a812c43af" |
| }, |
| { |
| "dataPath": "params_shard_6.bin", |
| "format": "raw-shard", |
| "nbytes": 25444352, |
| "records": [ |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.2.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.2.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 2101248 |
| }, |
| { |
| "name": "model.layers.2.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 5246976 |
| }, |
| { |
| "name": "model.layers.2.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5640192 |
| }, |
| { |
| "name": "model.layers.2.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7737344 |
| }, |
| { |
| "name": "model.layers.3.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 7999488 |
| }, |
| { |
| "name": "model.layers.3.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 8003584 |
| }, |
| { |
| "name": "model.layers.3.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 16392192 |
| }, |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 17440768 |
| }, |
| { |
| "name": "model.layers.3.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 19537920 |
| }, |
| { |
| "name": "model.layers.3.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 19542016 |
| }, |
| { |
| "name": "model.layers.3.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 22687744 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 23080960 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 25178112 |
| }, |
| { |
| "name": "model.layers.4.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 25440256 |
| } |
| ], |
| "md5sum": "9c1d4673e1bca543e3e801c9e5208640" |
| }, |
| { |
| "dataPath": "params_shard_7.bin", |
| "format": "raw-shard", |
| "nbytes": 31854592, |
| "records": [ |
| { |
| "name": "model.layers.4.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.4.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 8388608 |
| }, |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 9437184 |
| }, |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26214400 |
| }, |
| { |
| "name": "model.layers.4.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28311552 |
| }, |
| { |
| "name": "model.layers.4.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 28315648 |
| }, |
| { |
| "name": "model.layers.4.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 31461376 |
| } |
| ], |
| "md5sum": "9b11a54d902c8f27e5992fc98c50edfe" |
| }, |
| { |
| "dataPath": "params_shard_8.bin", |
| "format": "raw-shard", |
| "nbytes": 30679040, |
| "records": [ |
| { |
| "name": "model.layers.4.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.4.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.5.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2359296 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 2363392 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10752000 |
| }, |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 11800576 |
| }, |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 28577792 |
| }, |
| { |
| "name": "model.layers.5.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 30674944 |
| } |
| ], |
| "md5sum": "f264db50fdc16abe2e5b215740b80561" |
| }, |
| { |
| "dataPath": "params_shard_9.bin", |
| "format": "raw-shard", |
| "nbytes": 32116736, |
| "records": [ |
| { |
| "name": "model.layers.5.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.5.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.5.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.5.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 5636096 |
| }, |
| { |
| "name": "model.layers.6.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 5898240 |
| }, |
| { |
| "name": "model.layers.6.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 5902336 |
| }, |
| { |
| "name": "model.layers.6.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 14290944 |
| }, |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 15339520 |
| } |
| ], |
| "md5sum": "9f876b1f5066bc5b7e824467b3e84d75" |
| }, |
| { |
| "dataPath": "params_shard_10.bin", |
| "format": "raw-shard", |
| "nbytes": 16777216, |
| "records": [ |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "925bb8afdc06f942ffd65945f4870a17" |
| }, |
| { |
| "dataPath": "params_shard_11.bin", |
| "format": "raw-shard", |
| "nbytes": 25444352, |
| "records": [ |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.6.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.6.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 2101248 |
| }, |
| { |
| "name": "model.layers.6.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 5246976 |
| }, |
| { |
| "name": "model.layers.6.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5640192 |
| }, |
| { |
| "name": "model.layers.6.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7737344 |
| }, |
| { |
| "name": "model.layers.7.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 7999488 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 8003584 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 16392192 |
| }, |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 17440768 |
| }, |
| { |
| "name": "model.layers.7.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 19537920 |
| }, |
| { |
| "name": "model.layers.7.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 19542016 |
| }, |
| { |
| "name": "model.layers.7.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 22687744 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 23080960 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 25178112 |
| }, |
| { |
| "name": "model.layers.8.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 25440256 |
| } |
| ], |
| "md5sum": "c498ab27342dbf6b8c172f2cc4d76e9d" |
| }, |
| { |
| "dataPath": "params_shard_12.bin", |
| "format": "raw-shard", |
| "nbytes": 31854592, |
| "records": [ |
| { |
| "name": "model.layers.8.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.8.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 8388608 |
| }, |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 9437184 |
| }, |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26214400 |
| }, |
| { |
| "name": "model.layers.8.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28311552 |
| }, |
| { |
| "name": "model.layers.8.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 28315648 |
| }, |
| { |
| "name": "model.layers.8.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 31461376 |
| } |
| ], |
| "md5sum": "4571be920875718db296029a6fb5cf1d" |
| }, |
| { |
| "dataPath": "params_shard_13.bin", |
| "format": "raw-shard", |
| "nbytes": 30679040, |
| "records": [ |
| { |
| "name": "model.layers.8.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.8.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.9.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2359296 |
| }, |
| { |
| "name": "model.layers.9.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 2363392 |
| }, |
| { |
| "name": "model.layers.9.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10752000 |
| }, |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 11800576 |
| }, |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 28577792 |
| }, |
| { |
| "name": "model.layers.9.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 30674944 |
| } |
| ], |
| "md5sum": "b7d1ea446072aa0332c3ba4f07561c87" |
| }, |
| { |
| "dataPath": "params_shard_14.bin", |
| "format": "raw-shard", |
| "nbytes": 32116736, |
| "records": [ |
| { |
| "name": "model.layers.9.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.9.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.9.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.9.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 5636096 |
| }, |
| { |
| "name": "model.layers.10.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 5898240 |
| }, |
| { |
| "name": "model.layers.10.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 5902336 |
| }, |
| { |
| "name": "model.layers.10.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 14290944 |
| }, |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 15339520 |
| } |
| ], |
| "md5sum": "e143c828a701901a9d0b22dea6d2d570" |
| }, |
| { |
| "dataPath": "params_shard_15.bin", |
| "format": "raw-shard", |
| "nbytes": 16777216, |
| "records": [ |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "513a4806c0f7cb388b488a902eb61bd8" |
| }, |
| { |
| "dataPath": "params_shard_16.bin", |
| "format": "raw-shard", |
| "nbytes": 25444352, |
| "records": [ |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.10.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.10.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 2101248 |
| }, |
| { |
| "name": "model.layers.10.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 5246976 |
| }, |
| { |
| "name": "model.layers.10.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5640192 |
| }, |
| { |
| "name": "model.layers.10.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7737344 |
| }, |
| { |
| "name": "model.layers.11.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 7999488 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 8003584 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 16392192 |
| }, |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 17440768 |
| }, |
| { |
| "name": "model.layers.11.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 19537920 |
| }, |
| { |
| "name": "model.layers.11.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 19542016 |
| }, |
| { |
| "name": "model.layers.11.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 22687744 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 23080960 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 25178112 |
| }, |
| { |
| "name": "model.layers.12.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 25440256 |
| } |
| ], |
| "md5sum": "c7cc3267af48a73630d0ac72609c4e5e" |
| }, |
| { |
| "dataPath": "params_shard_17.bin", |
| "format": "raw-shard", |
| "nbytes": 31854592, |
| "records": [ |
| { |
| "name": "model.layers.12.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.12.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 8388608 |
| }, |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 9437184 |
| }, |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26214400 |
| }, |
| { |
| "name": "model.layers.12.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28311552 |
| }, |
| { |
| "name": "model.layers.12.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 28315648 |
| }, |
| { |
| "name": "model.layers.12.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 31461376 |
| } |
| ], |
| "md5sum": "58d7ff0d9edefbef420d80cb770a4ba7" |
| }, |
| { |
| "dataPath": "params_shard_18.bin", |
| "format": "raw-shard", |
| "nbytes": 30679040, |
| "records": [ |
| { |
| "name": "model.layers.12.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.12.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.13.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2359296 |
| }, |
| { |
| "name": "model.layers.13.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 2363392 |
| }, |
| { |
| "name": "model.layers.13.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10752000 |
| }, |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 11800576 |
| }, |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 28577792 |
| }, |
| { |
| "name": "model.layers.13.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 30674944 |
| } |
| ], |
| "md5sum": "841ed9a961bee09bf79f381545004e64" |
| }, |
| { |
| "dataPath": "params_shard_19.bin", |
| "format": "raw-shard", |
| "nbytes": 32116736, |
| "records": [ |
| { |
| "name": "model.layers.13.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.13.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.13.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.13.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 5636096 |
| }, |
| { |
| "name": "model.layers.14.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 5898240 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 5902336 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 14290944 |
| }, |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 15339520 |
| } |
| ], |
| "md5sum": "bf59448f9f8bf16dc48f89b4d5f17ff1" |
| }, |
| { |
| "dataPath": "params_shard_20.bin", |
| "format": "raw-shard", |
| "nbytes": 16777216, |
| "records": [ |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 16384, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 16777216, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "20291252b5e50485abe7d53716d97982" |
| }, |
| { |
| "dataPath": "params_shard_21.bin", |
| "format": "raw-shard", |
| "nbytes": 25444352, |
| "records": [ |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.14.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2097152 |
| }, |
| { |
| "name": "model.layers.14.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 2101248 |
| }, |
| { |
| "name": "model.layers.14.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 5246976 |
| }, |
| { |
| "name": "model.layers.14.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5640192 |
| }, |
| { |
| "name": "model.layers.14.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7737344 |
| }, |
| { |
| "name": "model.layers.15.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 7999488 |
| }, |
| { |
| "name": "model.layers.15.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8388608, |
| "byteOffset": 8003584 |
| }, |
| { |
| "name": "model.layers.15.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 16392192 |
| }, |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 16384, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 17440768 |
| }, |
| { |
| "name": "model.layers.15.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 19537920 |
| }, |
| { |
| "name": "model.layers.15.self_attn.qkv_proj.q_weight", |
| "shape": [ |
| 3072, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 19542016 |
| }, |
| { |
| "name": "model.layers.15.self_attn.qkv_proj.q_scale", |
| "shape": [ |
| 3072, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 22687744 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 23080960 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 25178112 |
| }, |
| { |
| "name": "model.norm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 25440256 |
| } |
| ], |
| "md5sum": "ddf57a84c42358e8c2c8e2778a24e5a3" |
| } |
| ] |
| } |
