| { | |
| "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": "529aed61a274c19024f5c9431c844191" | |
| }, | |
| { | |
| "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": "2d96a77f3cfae1681348920ced8a4eb3" | |
| }, | |
| { | |
| "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": "9d0308d4a05f2de6fad70696b440c121" | |
| }, | |
| { | |
| "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": "7d494db3d6a8878b7bb172141d2c199c" | |
| }, | |
| { | |
| "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": "e840ccb378123f884b610b7998317ee9" | |
| }, | |
| { | |
| "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": "74576718f84e42277b8f19de86c59d38" | |
| }, | |
| { | |
| "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": "4c9e508625af3d20843d2a2790c8aec3" | |
| }, | |
| { | |
| "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": "e0a5e676204344b3becc92ba4fc49483" | |
| }, | |
| { | |
| "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": "b54484ab4413dbc08fa159870d527d68" | |
| }, | |
| { | |
| "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": "439392412b181319d53217139d57a529" | |
| }, | |
| { | |
| "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": "8188330378814ad37db4cfdf00a11d95" | |
| }, | |
| { | |
| "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": "88116ac7734f2ac6b4aad2c585ccab24" | |
| }, | |
| { | |
| "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": "123ffa386f20421663a99f1033e5ff67" | |
| }, | |
| { | |
| "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": "a12c70a4f5af7b406b8f8fd9d6892ba8" | |
| }, | |
| { | |
| "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": "7e1ba19a0423a71138d171f8fb5cd6c4" | |
| }, | |
| { | |
| "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": "0a65248fcd87c0546ff9ce6de381c2aa" | |
| }, | |
| { | |
| "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": "3620069ece477b2428c167bff8c8bec0" | |
| }, | |
| { | |
| "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": "3533cabd22c9ccc6c3b9518d080f6fe3" | |
| }, | |
| { | |
| "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": "d0d5f3c421dc6a01f04100e60b9a7f99" | |
| }, | |
| { | |
| "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": "e29abdf82066a47e03d1723ffe9455f8" | |
| }, | |
| { | |
| "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": "bcfea4941432f0c90c83b73303fcc44d" | |
| }, | |
| { | |
| "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": "ddd77803cd6fe6cb534a0eafbc601517" | |
| }, | |
| { | |
| "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": "1886c5bf4a3708b2f7302e684b803ba9" | |
| }, | |
| { | |
| "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": "84934420226bf9efa26b8b92da7315ba" | |
| }, | |
| { | |
| "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": "c17d58a6907b1cb39e2dfa80dca9d6f5" | |
| }, | |
| { | |
| "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": "731905a27575670d4b05e8d8e0360005" | |
| }, | |
| { | |
| "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": "b18afcd171a835bdbf6c5b0b99817c05" | |
| }, | |
| { | |
| "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": "0e4c54c93e48cee927332b7ff179ceeb" | |
| }, | |
| { | |
| "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": "6a7c8b48ee29be223364e4d4eabc3365" | |
| }, | |
| { | |
| "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": "8e9ccfa6cdd42d1e1d2a55d6a25aed42" | |
| } | |
| ] | |
| } |