| { |
| "metadata": { |
| "ParamSize": 435, |
| "ParamBytes": 423938816.0, |
| "BitsPerParam": 4.507631543378053 |
| }, |
| "records": [ |
| { |
| "dataPath": "params_shard_0.bin", |
| "format": "raw-shard", |
| "nbytes": 127139840, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_weight", |
| "shape": [ |
| 248320, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 127139840, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "965719745e8bff8aee1dcdaaee99d0b9" |
| }, |
| { |
| "dataPath": "params_shard_1.bin", |
| "format": "raw-shard", |
| "nbytes": 32844416, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_scale", |
| "shape": [ |
| 248320, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 15892480, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.0.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 15892480 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 15894528 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 15894560 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 15943712 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 15943744 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 15951936 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 15952960 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 15961152 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 15962176 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 19107904 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 19501120 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 20549696 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 20680768 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 20681024 |
| }, |
| { |
| "name": "model.layers.0.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 21729600 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 21860672 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 23695680 |
| }, |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 23925056 |
| }, |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 27595072 |
| }, |
| { |
| "name": "model.layers.0.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 28053824 |
| }, |
| { |
| "name": "model.layers.1.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 28055872 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 28057920 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 28057952 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 28107104 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 28107136 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 28115328 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 28116352 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 28124544 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 28125568 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 31271296 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 31664512 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 32713088 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 32844160 |
| } |
| ], |
| "md5sum": "bace72e1b58ffe9a497e9b5a7eaffa6b" |
| }, |
| { |
| "dataPath": "params_shard_2.bin", |
| "format": "raw-shard", |
| "nbytes": 33473920, |
| "records": [ |
| { |
| "name": "model.layers.1.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.1.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 1048576 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 1179648 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 3014656 |
| }, |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 3244032 |
| }, |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 6914048 |
| }, |
| { |
| "name": "model.layers.1.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 7372800 |
| }, |
| { |
| "name": "model.layers.10.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 7374848 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 7376896 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 7376928 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 7426080 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 7426112 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 7434304 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 7435328 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 7443520 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 7444544 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 10590272 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10983488 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 12032064 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 12163136 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 12163392 |
| }, |
| { |
| "name": "model.layers.10.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 13211968 |
| }, |
| { |
| "name": "model.layers.10.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 13343040 |
| }, |
| { |
| "name": "model.layers.10.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 15178048 |
| }, |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 15407424 |
| }, |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 19077440 |
| }, |
| { |
| "name": "model.layers.10.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 19536192 |
| }, |
| { |
| "name": "model.layers.11.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 19538240 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 19540288 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 21375296 |
| }, |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 21604672 |
| }, |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 25274688 |
| }, |
| { |
| "name": "model.layers.11.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 25733440 |
| }, |
| { |
| "name": "model.layers.11.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 25735488 |
| }, |
| { |
| "name": "model.layers.11.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 25736000 |
| }, |
| { |
| "name": "model.layers.11.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 28357440 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 28685120 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 29733696 |
| }, |
| { |
| "name": "model.layers.11.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 29864768 |
| }, |
| { |
| "name": "model.layers.12.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 29865280 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 29867328 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 29867360 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 29916512 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 29916544 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 29924736 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 29925760 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 29933952 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 29934976 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 33080704 |
| } |
| ], |
| "md5sum": "4326282962d3cccd386c6ec6480dae26" |
| }, |
| { |
| "dataPath": "params_shard_3.bin", |
| "format": "raw-shard", |
| "nbytes": 32883584, |
| "records": [ |
| { |
| "name": "model.layers.12.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 1048576 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 1179648 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 1179904 |
| }, |
| { |
| "name": "model.layers.12.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 2228480 |
| }, |
| { |
| "name": "model.layers.12.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 2359552 |
| }, |
| { |
| "name": "model.layers.12.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 4194560 |
| }, |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 4423936 |
| }, |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 8093952 |
| }, |
| { |
| "name": "model.layers.12.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 8552704 |
| }, |
| { |
| "name": "model.layers.13.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 8554752 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 8556800 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 8556832 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 8605984 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 8606016 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 8614208 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 8615232 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 8623424 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 8624448 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 11770176 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 12163392 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 13211968 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 13343040 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 13343296 |
| }, |
| { |
| "name": "model.layers.13.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 14391872 |
| }, |
| { |
| "name": "model.layers.13.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 14522944 |
| }, |
| { |
| "name": "model.layers.13.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 16357952 |
| }, |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 16587328 |
| }, |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 20257344 |
| }, |
| { |
| "name": "model.layers.13.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 20716096 |
| }, |
| { |
| "name": "model.layers.14.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 20718144 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 20720192 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 20720224 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 20769376 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 20769408 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 20777600 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 20778624 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 20786816 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 20787840 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 23933568 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 24326784 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 25375360 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 25506432 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 25506688 |
| }, |
| { |
| "name": "model.layers.14.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 26555264 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 26686336 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 28521344 |
| }, |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 28750720 |
| }, |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 32420736 |
| }, |
| { |
| "name": "model.layers.14.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 32879488 |
| }, |
| { |
| "name": "model.layers.15.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 32881536 |
| } |
| ], |
| "md5sum": "11507c7398d6ba6237ff7698fd1ad943" |
| }, |
| { |
| "dataPath": "params_shard_4.bin", |
| "format": "raw-shard", |
| "nbytes": 30520960, |
| "records": [ |
| { |
| "name": "model.layers.15.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.15.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 1835008 |
| }, |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 2064384 |
| }, |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 5734400 |
| }, |
| { |
| "name": "model.layers.15.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 6193152 |
| }, |
| { |
| "name": "model.layers.15.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 6195200 |
| }, |
| { |
| "name": "model.layers.15.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 6195712 |
| }, |
| { |
| "name": "model.layers.15.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 8817152 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 9144832 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 10193408 |
| }, |
| { |
| "name": "model.layers.15.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 10324480 |
| }, |
| { |
| "name": "model.layers.16.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 10324992 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 10327040 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 10327072 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 10376224 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 10376256 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 10384448 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 10385472 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 10393664 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 10394688 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 13540416 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 13933632 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 14982208 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 15113280 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 15113536 |
| }, |
| { |
| "name": "model.layers.16.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 16162112 |
| }, |
| { |
| "name": "model.layers.16.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 16293184 |
| }, |
| { |
| "name": "model.layers.16.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 18128192 |
| }, |
| { |
| "name": "model.layers.16.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 18357568 |
| }, |
| { |
| "name": "model.layers.16.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 22027584 |
| }, |
| { |
| "name": "model.layers.16.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 22486336 |
| }, |
| { |
| "name": "model.layers.17.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 22488384 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 22490432 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 22490464 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 22539616 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 22539648 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 22547840 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 22548864 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 22557056 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 22558080 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 25703808 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 26097024 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 27145600 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 27276672 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 27276928 |
| }, |
| { |
| "name": "model.layers.17.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 28325504 |
| }, |
| { |
| "name": "model.layers.17.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 28456576 |
| }, |
| { |
| "name": "model.layers.17.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 30291584 |
| } |
| ], |
| "md5sum": "d98f48b1c21e58e1f0e80cc08fd7deed" |
| }, |
| { |
| "dataPath": "params_shard_5.bin", |
| "format": "raw-shard", |
| "nbytes": 32589440, |
| "records": [ |
| { |
| "name": "model.layers.17.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.17.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 3670016 |
| }, |
| { |
| "name": "model.layers.17.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 4128768 |
| }, |
| { |
| "name": "model.layers.18.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 4130816 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 4132864 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 4132896 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 4182048 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 4182080 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 4190272 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 4191296 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 4199488 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 4200512 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 7346240 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 7739456 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 8788032 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 8919104 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 8919360 |
| }, |
| { |
| "name": "model.layers.18.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 9967936 |
| }, |
| { |
| "name": "model.layers.18.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 10099008 |
| }, |
| { |
| "name": "model.layers.18.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 11934016 |
| }, |
| { |
| "name": "model.layers.18.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 12163392 |
| }, |
| { |
| "name": "model.layers.18.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 15833408 |
| }, |
| { |
| "name": "model.layers.18.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 16292160 |
| }, |
| { |
| "name": "model.layers.19.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 16294208 |
| }, |
| { |
| "name": "model.layers.19.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 16296256 |
| }, |
| { |
| "name": "model.layers.19.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 18131264 |
| }, |
| { |
| "name": "model.layers.19.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 18360640 |
| }, |
| { |
| "name": "model.layers.19.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 22030656 |
| }, |
| { |
| "name": "model.layers.19.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 22489408 |
| }, |
| { |
| "name": "model.layers.19.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 22491456 |
| }, |
| { |
| "name": "model.layers.19.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 22491968 |
| }, |
| { |
| "name": "model.layers.19.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 25113408 |
| }, |
| { |
| "name": "model.layers.19.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 25441088 |
| }, |
| { |
| "name": "model.layers.19.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 26489664 |
| }, |
| { |
| "name": "model.layers.19.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 26620736 |
| }, |
| { |
| "name": "model.layers.2.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 26621248 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 26623296 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 26623328 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 26672480 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 26672512 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 26680704 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 26681728 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 26689920 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 26690944 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 29836672 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 30229888 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 31278464 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 31409536 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 31409792 |
| }, |
| { |
| "name": "model.layers.2.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 32458368 |
| } |
| ], |
| "md5sum": "2f15bf82023905de99ee433bc410b6cd" |
| }, |
| { |
| "dataPath": "params_shard_6.bin", |
| "format": "raw-shard", |
| "nbytes": 30591680, |
| "records": [ |
| { |
| "name": "model.layers.2.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.2.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 1835008 |
| }, |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 2064384 |
| }, |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 5734400 |
| }, |
| { |
| "name": "model.layers.2.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 6193152 |
| }, |
| { |
| "name": "model.layers.20.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 6195200 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 6197248 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 6197280 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 6246432 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 6246464 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 6254656 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 6255680 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 6263872 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 6264896 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 9410624 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 9803840 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 10852416 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 10983488 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 10983744 |
| }, |
| { |
| "name": "model.layers.20.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 12032320 |
| }, |
| { |
| "name": "model.layers.20.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 12163392 |
| }, |
| { |
| "name": "model.layers.20.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 13998400 |
| }, |
| { |
| "name": "model.layers.20.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 14227776 |
| }, |
| { |
| "name": "model.layers.20.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 17897792 |
| }, |
| { |
| "name": "model.layers.20.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 18356544 |
| }, |
| { |
| "name": "model.layers.21.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 18358592 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 18360640 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 18360672 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 18409824 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 18409856 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 18418048 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 18419072 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 18427264 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 18428288 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 21574016 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 21967232 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 23015808 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 23146880 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 23147136 |
| }, |
| { |
| "name": "model.layers.21.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 24195712 |
| }, |
| { |
| "name": "model.layers.21.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 24326784 |
| }, |
| { |
| "name": "model.layers.21.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 26161792 |
| }, |
| { |
| "name": "model.layers.21.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 26391168 |
| }, |
| { |
| "name": "model.layers.21.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 30061184 |
| }, |
| { |
| "name": "model.layers.21.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 30519936 |
| }, |
| { |
| "name": "model.layers.22.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 30521984 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 30524032 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 30524064 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 30573216 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 30573248 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 30581440 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 30582464 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 30590656 |
| } |
| ], |
| "md5sum": "36b6f955709784df769c59208b708993" |
| }, |
| { |
| "dataPath": "params_shard_7.bin", |
| "format": "raw-shard", |
| "nbytes": 32817472, |
| "records": [ |
| { |
| "name": "model.layers.22.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 4587520 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 4718592 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 4718848 |
| }, |
| { |
| "name": "model.layers.22.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 5767424 |
| }, |
| { |
| "name": "model.layers.22.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 5898496 |
| }, |
| { |
| "name": "model.layers.22.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 7733504 |
| }, |
| { |
| "name": "model.layers.22.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 7962880 |
| }, |
| { |
| "name": "model.layers.22.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 11632896 |
| }, |
| { |
| "name": "model.layers.22.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 12091648 |
| }, |
| { |
| "name": "model.layers.23.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 12093696 |
| }, |
| { |
| "name": "model.layers.23.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 12095744 |
| }, |
| { |
| "name": "model.layers.23.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 13930752 |
| }, |
| { |
| "name": "model.layers.23.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 14160128 |
| }, |
| { |
| "name": "model.layers.23.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 17830144 |
| }, |
| { |
| "name": "model.layers.23.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 18288896 |
| }, |
| { |
| "name": "model.layers.23.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 18290944 |
| }, |
| { |
| "name": "model.layers.23.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 18291456 |
| }, |
| { |
| "name": "model.layers.23.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 20912896 |
| }, |
| { |
| "name": "model.layers.23.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 21240576 |
| }, |
| { |
| "name": "model.layers.23.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 22289152 |
| }, |
| { |
| "name": "model.layers.23.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 22420224 |
| }, |
| { |
| "name": "model.layers.3.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 22420736 |
| }, |
| { |
| "name": "model.layers.3.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 22422784 |
| }, |
| { |
| "name": "model.layers.3.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 24257792 |
| }, |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 24487168 |
| }, |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 28157184 |
| }, |
| { |
| "name": "model.layers.3.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 28615936 |
| }, |
| { |
| "name": "model.layers.3.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 28617984 |
| }, |
| { |
| "name": "model.layers.3.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 28618496 |
| }, |
| { |
| "name": "model.layers.3.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 31239936 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 31567616 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 32616192 |
| }, |
| { |
| "name": "model.layers.3.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 32747264 |
| }, |
| { |
| "name": "model.layers.4.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 32747776 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 32749824 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 32749856 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 32799008 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 32799040 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 32807232 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 32808256 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 32816448 |
| } |
| ], |
| "md5sum": "e5c3bb181a1b2410f00104e82a97095b" |
| }, |
| { |
| "dataPath": "params_shard_8.bin", |
| "format": "raw-shard", |
| "nbytes": 32289664, |
| "records": [ |
| { |
| "name": "model.layers.4.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 3145728 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 3538944 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 4587520 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 4718592 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 4718848 |
| }, |
| { |
| "name": "model.layers.4.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 5767424 |
| }, |
| { |
| "name": "model.layers.4.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 5898496 |
| }, |
| { |
| "name": "model.layers.4.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 7733504 |
| }, |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 7962880 |
| }, |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 11632896 |
| }, |
| { |
| "name": "model.layers.4.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 12091648 |
| }, |
| { |
| "name": "model.layers.5.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 12093696 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 12095744 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 12095776 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 12144928 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 12144960 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 12153152 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 12154176 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 12162368 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 12163392 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 15309120 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 15702336 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 16750912 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 16881984 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 16882240 |
| }, |
| { |
| "name": "model.layers.5.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 17930816 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 18061888 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 19896896 |
| }, |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 20126272 |
| }, |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 23796288 |
| }, |
| { |
| "name": "model.layers.5.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 24255040 |
| }, |
| { |
| "name": "model.layers.6.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 24257088 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 24259136 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 24259168 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 24308320 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 24308352 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 24316544 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 24317568 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 24325760 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 24326784 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 27472512 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 27865728 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 28914304 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 29045376 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 29045632 |
| }, |
| { |
| "name": "model.layers.6.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 30094208 |
| }, |
| { |
| "name": "model.layers.6.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 30225280 |
| }, |
| { |
| "name": "model.layers.6.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 32060288 |
| } |
| ], |
| "md5sum": "9062a4946203bfdf30f6a07a1cc91692" |
| }, |
| { |
| "dataPath": "params_shard_9.bin", |
| "format": "raw-shard", |
| "nbytes": 32589440, |
| "records": [ |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 3670016 |
| }, |
| { |
| "name": "model.layers.6.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 4128768 |
| }, |
| { |
| "name": "model.layers.7.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 4130816 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 4132864 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 5967872 |
| }, |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 6197248 |
| }, |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 9867264 |
| }, |
| { |
| "name": "model.layers.7.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 10326016 |
| }, |
| { |
| "name": "model.layers.7.self_attn.k_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 10328064 |
| }, |
| { |
| "name": "model.layers.7.self_attn.c_attn.q_weight", |
| "shape": [ |
| 5120, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 10328576 |
| }, |
| { |
| "name": "model.layers.7.self_attn.c_attn.q_scale", |
| "shape": [ |
| 5120, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 12950016 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 13277696 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 14326272 |
| }, |
| { |
| "name": "model.layers.7.self_attn.q_norm.weight", |
| "shape": [ |
| 256 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 512, |
| "byteOffset": 14457344 |
| }, |
| { |
| "name": "model.layers.8.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 14457856 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 14459904 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 14459936 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 14509088 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 14509120 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 14517312 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 14518336 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 14526528 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 14527552 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 17673280 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 18066496 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 19115072 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 19246144 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 19246400 |
| }, |
| { |
| "name": "model.layers.8.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 20294976 |
| }, |
| { |
| "name": "model.layers.8.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 20426048 |
| }, |
| { |
| "name": "model.layers.8.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 22261056 |
| }, |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 22490432 |
| }, |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 26160448 |
| }, |
| { |
| "name": "model.layers.8.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 26619200 |
| }, |
| { |
| "name": "model.layers.9.input_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 26621248 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.A_log", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 26623296 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.conv1d_weight", |
| "shape": [ |
| 6144, |
| 1, |
| 4 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 49152, |
| "byteOffset": 26623328 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.dt_bias", |
| "shape": [ |
| 16 |
| ], |
| "dtype": "float32", |
| "format": "f32-to-bf16", |
| "nbytes": 32, |
| "byteOffset": 26672480 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_a.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 26672512 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_a.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 26680704 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_b.q_weight", |
| "shape": [ |
| 16, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 8192, |
| "byteOffset": 26681728 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_b.q_scale", |
| "shape": [ |
| 16, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1024, |
| "byteOffset": 26689920 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_qkv.q_weight", |
| "shape": [ |
| 6144, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3145728, |
| "byteOffset": 26690944 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_qkv.q_scale", |
| "shape": [ |
| 6144, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 393216, |
| "byteOffset": 29836672 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_z.q_weight", |
| "shape": [ |
| 2048, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 30229888 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.in_proj_z.q_scale", |
| "shape": [ |
| 2048, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 31278464 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.norm.weight", |
| "shape": [ |
| 128 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 256, |
| "byteOffset": 31409536 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.out_proj.q_weight", |
| "shape": [ |
| 1024, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1048576, |
| "byteOffset": 31409792 |
| }, |
| { |
| "name": "model.layers.9.linear_attn.out_proj.q_scale", |
| "shape": [ |
| 1024, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 131072, |
| "byteOffset": 32458368 |
| } |
| ], |
| "md5sum": "3ce64a1408d65588ea658122f93b94fe" |
| }, |
| { |
| "dataPath": "params_shard_10.bin", |
| "format": "raw-shard", |
| "nbytes": 6197248, |
| "records": [ |
| { |
| "name": "model.layers.9.mlp.down_proj.q_weight", |
| "shape": [ |
| 1024, |
| 448 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 1835008, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.9.mlp.down_proj.q_scale", |
| "shape": [ |
| 1024, |
| 112 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 229376, |
| "byteOffset": 1835008 |
| }, |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 7168, |
| 128 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 3670016, |
| "byteOffset": 2064384 |
| }, |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 7168, |
| 32 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 458752, |
| "byteOffset": 5734400 |
| }, |
| { |
| "name": "model.layers.9.post_attention_layernorm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 6193152 |
| }, |
| { |
| "name": "model.norm.weight", |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2048, |
| "byteOffset": 6195200 |
| } |
| ], |
| "md5sum": "d619f57e9a9104b6683591eb1e5826bf" |
| } |
| ] |
| } |