| { |
| "metadata": { |
| "ParamSize": 399, |
| "ParamBytes": 1736187904.0, |
| "BitsPerParam": 4.500900580433049 |
| }, |
| "records": [ |
| { |
| "dataPath": "params_shard_0.bin", |
| "format": "raw-shard", |
| "nbytes": 155582464, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_weight", |
| "shape": [ |
| 151936, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 155582464, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "068d00b27019e8607e13f648c3473040" |
| }, |
| { |
| "dataPath": "params_shard_1.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "9cc0b6a29a6acec94b0d7c98d0b61b2b" |
| }, |
| { |
| "dataPath": "params_shard_2.bin", |
| "format": "raw-shard", |
| "nbytes": 32133120, |
| "records": [ |
| { |
| "name": "model.embed_tokens.q_scale", |
| "shape": [ |
| 151936, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 19447808, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.0.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 19447808 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 19451904 |
| }, |
| { |
| "name": "model.layers.0.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 30724096 |
| } |
| ], |
| "md5sum": "f9215e85504ba381a274072238555a1c" |
| }, |
| { |
| "dataPath": "params_shard_3.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "e09d9ee14dea76e96bcd08517252d23d" |
| }, |
| { |
| "dataPath": "params_shard_4.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.0.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.0.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.0.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.0.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.0.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.0.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.0.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.1.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.1.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.1.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.1.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.1.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.1.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.1.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.1.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.1.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.10.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "63e77073191b3efb4c6f82659aa6e0a8" |
| }, |
| { |
| "dataPath": "params_shard_5.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "bbdb9863e4a51b2424a643047926d235" |
| }, |
| { |
| "dataPath": "params_shard_6.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "274ca7894b6925f8a331ff0001d26dec" |
| }, |
| { |
| "dataPath": "params_shard_7.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.10.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.10.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.10.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.10.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.10.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.10.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.10.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.10.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.10.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.11.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.11.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "9c2c6fffa0bba59a7ea28ae50d36ade9" |
| }, |
| { |
| "dataPath": "params_shard_8.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "6296eb083558d03b66fe2637d4b5db6d" |
| }, |
| { |
| "dataPath": "params_shard_9.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.11.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.11.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.11.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.11.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.11.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.11.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.12.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.12.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.12.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.12.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.12.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.12.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.12.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.12.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.12.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.12.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.13.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "694a15ee4c20738ab7bc1ff388812bba" |
| }, |
| { |
| "dataPath": "params_shard_10.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "52397a241400ad47dcdf4e5c47528494" |
| }, |
| { |
| "dataPath": "params_shard_11.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "aaee11b77c3c1d77ae3861f014d73d99" |
| }, |
| { |
| "dataPath": "params_shard_12.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.13.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.13.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.13.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.13.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.13.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.13.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.13.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.13.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.13.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.14.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.14.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "1e5c613c8024d919f6a7c8413224dc48" |
| }, |
| { |
| "dataPath": "params_shard_13.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "b9f3417a77496ff958e60a845089cbad" |
| }, |
| { |
| "dataPath": "params_shard_14.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.14.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.14.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.14.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.14.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.14.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.14.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.14.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.15.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.15.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.15.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.15.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.15.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.15.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.15.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.15.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.15.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.16.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "5ae2583d8ddd2d455222a04abff4ede1" |
| }, |
| { |
| "dataPath": "params_shard_15.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.16.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "32c539995b8c9b8b8f0e43da6ac167d7" |
| }, |
| { |
| "dataPath": "params_shard_16.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.17.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "59c950acdfb9e8d4c00e6d062f2b312d" |
| }, |
| { |
| "dataPath": "params_shard_17.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.16.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.16.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.16.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.16.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.16.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.16.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.16.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.16.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.16.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.17.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.17.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.17.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "22b3106179876d206ced7186a58a3834" |
| }, |
| { |
| "dataPath": "params_shard_18.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.18.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "a3be926a5568d88f3828a67a776fd7c7" |
| }, |
| { |
| "dataPath": "params_shard_19.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.17.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.17.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.17.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.17.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.17.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.17.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.17.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.18.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.18.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.18.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.18.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.18.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.18.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.18.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.18.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.18.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.18.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.19.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "be5894e5177bb2f077968333fd27b479" |
| }, |
| { |
| "dataPath": "params_shard_20.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.19.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "a370cf21cedeebcb98f80ff204877e86" |
| }, |
| { |
| "dataPath": "params_shard_21.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "ab4c0a39ab40dad0bc1ba23166cae4db" |
| }, |
| { |
| "dataPath": "params_shard_22.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.19.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.19.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.19.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.19.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.19.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.19.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.19.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.19.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.19.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.2.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.2.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.2.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "3ec44f377fbbdb2289e06b8cc24b3a44" |
| }, |
| { |
| "dataPath": "params_shard_23.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.20.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "fa444b99e483ecf9c424aba81ef4cc70" |
| }, |
| { |
| "dataPath": "params_shard_24.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.2.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.2.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.2.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.2.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.2.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.2.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.2.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.20.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.20.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.20.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.20.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.20.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.20.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.20.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.20.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.20.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.20.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.21.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "899be24e42b00cd45afe95643e8cfd33" |
| }, |
| { |
| "dataPath": "params_shard_25.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.21.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "7aee547ccbcaa2b5c0e72712cf36a40b" |
| }, |
| { |
| "dataPath": "params_shard_26.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.22.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "61d9d1d43ad582ad59271a1fa9e4bde0" |
| }, |
| { |
| "dataPath": "params_shard_27.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.21.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.21.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.21.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.21.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.21.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.21.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.21.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.21.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.21.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.22.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.22.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.22.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "a4f2f7e5edead35af9b5732b068e4707" |
| }, |
| { |
| "dataPath": "params_shard_28.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.23.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "adb421c289d2ff8c8a8dbad8a28b4f33" |
| }, |
| { |
| "dataPath": "params_shard_29.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.22.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.22.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.22.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.22.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.22.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.22.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.22.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.23.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.23.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.23.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.23.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.23.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.23.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.23.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.23.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.23.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.23.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.24.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "b22df9456c709146b334d6347ceb65a7" |
| }, |
| { |
| "dataPath": "params_shard_30.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.24.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "5f220203c1be49dbf71664ff61acdc90" |
| }, |
| { |
| "dataPath": "params_shard_31.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.25.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "10aca4f9d12cb2b56026f2fbe3ec498e" |
| }, |
| { |
| "dataPath": "params_shard_32.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.24.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.24.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.24.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.24.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.24.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.24.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.24.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.24.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.24.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.25.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.25.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.25.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "90e73a8c38df821e93c6f3c0cbe6ec9e" |
| }, |
| { |
| "dataPath": "params_shard_33.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.26.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "5dee4d360086abe773f9b8bc1d223d8f" |
| }, |
| { |
| "dataPath": "params_shard_34.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.25.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.25.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.25.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.25.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.25.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.25.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.25.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.26.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.26.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.26.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.26.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.26.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.26.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.26.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.26.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.26.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.26.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.27.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "7ec1d2e6e4a53bd2daa39a8a90b329de" |
| }, |
| { |
| "dataPath": "params_shard_35.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.27.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "89495cee431cd1278af10e469bd52897" |
| }, |
| { |
| "dataPath": "params_shard_36.bin", |
| "format": "raw-shard", |
| "nbytes": 26134528, |
| "records": [ |
| { |
| "name": "model.layers.27.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.27.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.27.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.27.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.27.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.27.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.27.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.27.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.27.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.28.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.28.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 20822016 |
| }, |
| { |
| "name": "model.layers.28.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 23443456 |
| }, |
| { |
| "name": "model.layers.28.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 23771136 |
| }, |
| { |
| "name": "model.layers.28.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 25868288 |
| }, |
| { |
| "name": "model.layers.3.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 26130432 |
| } |
| ], |
| "md5sum": "f9d706fe46d1f14ba4d77ae828067cee" |
| }, |
| { |
| "dataPath": "params_shard_37.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "1c3d94427ead25f65f2097cbc53204a1" |
| }, |
| { |
| "dataPath": "params_shard_38.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "130f6dad7d949e59aaa6ba0a7e115b82" |
| }, |
| { |
| "dataPath": "params_shard_39.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.3.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.3.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.3.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.3.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.3.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.3.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.3.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.3.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.4.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.4.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.4.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "b4b5250ab7163cde65747db61f299e38" |
| }, |
| { |
| "dataPath": "params_shard_40.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "17ae822a3b82c3509a1c436585f72c49" |
| }, |
| { |
| "dataPath": "params_shard_41.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.4.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.4.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.4.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.4.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.4.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.4.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.4.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.5.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.5.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.5.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.5.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.5.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.5.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.5.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.5.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.5.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.6.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "80af53b74928234573b89a8ef942c0bf" |
| }, |
| { |
| "dataPath": "params_shard_42.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "cdfd784d692d2aaffb274236df08a886" |
| }, |
| { |
| "dataPath": "params_shard_43.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "4b17ecd55a9841fcd17cba9218054c16" |
| }, |
| { |
| "dataPath": "params_shard_44.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.6.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.6.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.6.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.6.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.6.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.6.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.6.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.6.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.6.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.7.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.7.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "054e3d420572375bbc83199b1f38a09c" |
| }, |
| { |
| "dataPath": "params_shard_45.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "dccceebc4115ee248189b8d8a0569f6c" |
| }, |
| { |
| "dataPath": "params_shard_46.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.7.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.7.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.7.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.7.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.7.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.7.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.8.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.8.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.8.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.8.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.8.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.8.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.8.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.8.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.8.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.8.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.9.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "8b82f71a1260eef97a3f8f64a9248188" |
| }, |
| { |
| "dataPath": "params_shard_47.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "718b80f7b5bd994b0a766cbe8d17464a" |
| }, |
| { |
| "dataPath": "params_shard_48.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.28.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "511691076c0f0deec567e17a4c957ceb" |
| }, |
| { |
| "dataPath": "params_shard_49.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.9.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.9.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.9.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.9.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.9.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.9.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.9.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.9.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.9.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.28.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.28.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.28.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "ee03e389dbff7ef232e4974590ef6103" |
| }, |
| { |
| "dataPath": "params_shard_50.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.29.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "eefb0aeb62699c04c50510d4ea96d39f" |
| }, |
| { |
| "dataPath": "params_shard_51.bin", |
| "format": "raw-shard", |
| "nbytes": 23647232, |
| "records": [ |
| { |
| "name": "model.layers.28.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.28.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.29.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.29.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 2826240 |
| }, |
| { |
| "name": "model.layers.29.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 14098432 |
| }, |
| { |
| "name": "model.layers.29.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 15507456 |
| }, |
| { |
| "name": "model.layers.29.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 18325504 |
| }, |
| { |
| "name": "model.layers.29.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 18329600 |
| }, |
| { |
| "name": "model.layers.29.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 18334720 |
| }, |
| { |
| "name": "model.layers.29.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 20956160 |
| }, |
| { |
| "name": "model.layers.29.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 21283840 |
| }, |
| { |
| "name": "model.layers.29.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 23380992 |
| }, |
| { |
| "name": "model.layers.30.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23643136 |
| } |
| ], |
| "md5sum": "175600de1014580b5be147663e89d2cc" |
| }, |
| { |
| "dataPath": "params_shard_52.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.30.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "3660e549041ce4e435e9493ab5b68aba" |
| }, |
| { |
| "dataPath": "params_shard_53.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.31.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "fdc1f608a0ceee55d4462d42d6726e65" |
| }, |
| { |
| "dataPath": "params_shard_54.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.30.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.30.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.30.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.30.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.30.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.30.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.30.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.30.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.30.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.31.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.31.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.31.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "ffeeb4eee5272448a0820f3bbb8c9827" |
| }, |
| { |
| "dataPath": "params_shard_55.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.32.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "de48d4fb165b2ec7b72e71687b4371e7" |
| }, |
| { |
| "dataPath": "params_shard_56.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.31.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.31.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.31.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.31.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.31.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.31.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.31.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.32.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.32.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.32.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.32.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.32.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.32.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.32.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.32.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.32.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.32.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.layers.33.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "7aaa27dd9977887925e43a9c818256dd" |
| }, |
| { |
| "dataPath": "params_shard_57.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.33.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "c5e4efa6bb73da52697ff0d59cfcfd65" |
| }, |
| { |
| "dataPath": "params_shard_58.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.34.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "5845dcd27784b63488dd0de56eeedff3" |
| }, |
| { |
| "dataPath": "params_shard_59.bin", |
| "format": "raw-shard", |
| "nbytes": 33502208, |
| "records": [ |
| { |
| "name": "model.layers.33.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.33.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 11272192 |
| }, |
| { |
| "name": "model.layers.33.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 12681216 |
| }, |
| { |
| "name": "model.layers.33.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 15499264 |
| }, |
| { |
| "name": "model.layers.33.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 15503360 |
| }, |
| { |
| "name": "model.layers.33.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 15508480 |
| }, |
| { |
| "name": "model.layers.33.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 18129920 |
| }, |
| { |
| "name": "model.layers.33.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 18457600 |
| }, |
| { |
| "name": "model.layers.33.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 20554752 |
| }, |
| { |
| "name": "model.layers.34.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 20816896 |
| }, |
| { |
| "name": "model.layers.34.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.34.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 32093184 |
| } |
| ], |
| "md5sum": "bdcd2f569f13e74a054cc1ca66ad1f10" |
| }, |
| { |
| "dataPath": "params_shard_60.bin", |
| "format": "raw-shard", |
| "nbytes": 22544384, |
| "records": [ |
| { |
| "name": "model.layers.35.mlp.gate_up_proj.q_weight", |
| "shape": [ |
| 22016, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 22544384, |
| "byteOffset": 0 |
| } |
| ], |
| "md5sum": "3057bd2ce49afb0b41298b69e02dee6d" |
| }, |
| { |
| "dataPath": "params_shard_61.bin", |
| "format": "raw-shard", |
| "nbytes": 28960768, |
| "records": [ |
| { |
| "name": "model.layers.34.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 0 |
| }, |
| { |
| "name": "model.layers.34.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 2818048 |
| }, |
| { |
| "name": "model.layers.34.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 2822144 |
| }, |
| { |
| "name": "model.layers.34.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 2827264 |
| }, |
| { |
| "name": "model.layers.34.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 5448704 |
| }, |
| { |
| "name": "model.layers.34.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 5776384 |
| }, |
| { |
| "name": "model.layers.34.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 7873536 |
| }, |
| { |
| "name": "model.layers.35.input_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 8135680 |
| }, |
| { |
| "name": "model.layers.35.mlp.down_proj.q_weight", |
| "shape": [ |
| 2048, |
| 1376 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 11272192, |
| "byteOffset": 8139776 |
| }, |
| { |
| "name": "model.layers.35.mlp.down_proj.q_scale", |
| "shape": [ |
| 2048, |
| 344 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 1409024, |
| "byteOffset": 19411968 |
| }, |
| { |
| "name": "model.layers.35.mlp.gate_up_proj.q_scale", |
| "shape": [ |
| 22016, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 2818048, |
| "byteOffset": 20820992 |
| }, |
| { |
| "name": "model.layers.35.post_attention_layernorm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 23639040 |
| }, |
| { |
| "name": "model.layers.35.self_attn.c_attn.bias", |
| "shape": [ |
| 2560 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 5120, |
| "byteOffset": 23643136 |
| }, |
| { |
| "name": "model.layers.35.self_attn.c_attn.q_weight", |
| "shape": [ |
| 2560, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2621440, |
| "byteOffset": 23648256 |
| }, |
| { |
| "name": "model.layers.35.self_attn.c_attn.q_scale", |
| "shape": [ |
| 2560, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 327680, |
| "byteOffset": 26269696 |
| }, |
| { |
| "name": "model.layers.35.self_attn.o_proj.q_weight", |
| "shape": [ |
| 2048, |
| 256 |
| ], |
| "dtype": "uint32", |
| "format": "f32-to-bf16", |
| "nbytes": 2097152, |
| "byteOffset": 26597376 |
| }, |
| { |
| "name": "model.layers.35.self_attn.o_proj.q_scale", |
| "shape": [ |
| 2048, |
| 64 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 262144, |
| "byteOffset": 28694528 |
| }, |
| { |
| "name": "model.norm.weight", |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "float16", |
| "format": "f32-to-bf16", |
| "nbytes": 4096, |
| "byteOffset": 28956672 |
| } |
| ], |
| "md5sum": "5a3e2d42df5a25cceb1a27dc2ac7ba16" |
| } |
| ] |
| } |
