diff --git "a/ndarray-cache.json" "b/ndarray-cache.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache.json" @@ -0,0 +1,4785 @@ +{ + "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": "a8f329f6ab0b7bc4d3b142d0f990cc24" + }, + { + "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": "11621cd4b1fbd1a6bf28fa42866098cb" + }, + { + "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": "0ed8548cb381dec97a27922669579e92" + }, + { + "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": "18a0910b4e572b42c772684643d8879a" + }, + { + "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": "f9a226978881ae7ba6458975dd83628e" + }, + { + "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": "b1a02fe0851610a0a181efcae641d1a0" + }, + { + "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": "c6b9efc12099b53746e9a3fb22109c90" + }, + { + "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": "43f810349fc34e73d49773a43b2691be" + }, + { + "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": "c1c0edd7bae6cc9801a19f813f0a875b" + }, + { + "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": "06f69167e277ec320c3cafef5d258a94" + }, + { + "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": "6c326b56d9e0aa2c1f94569d9807e8da" + }, + { + "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": "eb1ee7869d5aa07d2ef427eec7cffadf" + }, + { + "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": "9d72412a3a00300ce4ac2c543fbe912f" + }, + { + "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": "c2c410aae1e01fab011410aaaa5b58d2" + }, + { + "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": "1a2b98459c3177cd2ab18b5c0242e975" + }, + { + "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": "50fe624cc1f74eb46e2af5fb3e07c673" + }, + { + "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": "dd0d4af95ba1a50bcbb25647bd4f0e1a" + }, + { + "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": "902f9bdef24591ae5daa11e674598470" + }, + { + "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": "a11e2c9b321ec0a43d1f931023611e23" + }, + { + "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": "465f73217a90b4e70de377d28356174f" + }, + { + "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": "dc3dee7770afcc741de2e14a6485c176" + }, + { + "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": "1e91ae347581b376d1b8fcc57efe907e" + }, + { + "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": "82b4c0c1091a6c10cfa47711e4f5427c" + }, + { + "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": "d4c090020fb9b076d3b822d5272d7c95" + }, + { + "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": "835f2859668a9a0ed975bb39e374b862" + }, + { + "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": "744ddbc88bb00ee30cf632ced4127195" + }, + { + "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": "52a8cd27173dd6f7d92633ddce93ea65" + }, + { + "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": "15591d7aa011814c02652c1ffd978cd6" + }, + { + "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": "51bfa386319627c050432bee01093cff" + }, + { + "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": "e421ab69552076c3757c3c4f06b9e777" + }, + { + "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": "1f3dc306f78d06ba50f70e0b4a7d2382" + }, + { + "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": "95296fcd0bd9d11efb19a2fa9dc615a8" + }, + { + "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": "4a64864b48b84a48480fc5470393781d" + }, + { + "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": "6a6beb18e7d0d836c68a1bbd716472dc" + }, + { + "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": "656e009e42fa4f192040154e5bda2279" + }, + { + "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": "d6346d860e76804316b94b1c4fab1e2c" + }, + { + "dataPath": "params_shard_36.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": "d2ed695a602c395fcaf935bcb446a893" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 33502208, + "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.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": "b976732eea5181a579d58505aac4701e" + }, + { + "dataPath": "params_shard_38.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": "b9639b7f27e9f5897613e8016da0d0f2" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 28960768, + "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.28.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.28.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.28.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.29.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.29.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.29.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "d8aa855ba014d321df51b8745c61c9b5" + }, + { + "dataPath": "params_shard_40.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": "78dd38a3d37a287219193ba99154668e" + }, + { + "dataPath": "params_shard_41.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": "32237f73544e0f5571e12acf488c32d8" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 33498112, + "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.30.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20816896 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32089088 + } + ], + "md5sum": "cacb240673dcc11e0201d1943cc4e881" + }, + { + "dataPath": "params_shard_43.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": "f8a54c6535ad0394d0bfb6e151326edc" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 28956672, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.30.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2818048 + }, + { + "name": "model.layers.30.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2823168 + }, + { + "name": "model.layers.30.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5444608 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5772288 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7869440 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8131584 + }, + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8135680 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19407872 + }, + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20816896 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23634944 + }, + { + "name": "model.layers.4.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23639040 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23644160 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26265600 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26593280 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28690432 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28952576 + } + ], + "md5sum": "60340ed231c87302578920db11ba987b" + }, + { + "dataPath": "params_shard_45.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": "785188cbd5bb969efec41aba675e9d5a" + }, + { + "dataPath": "params_shard_46.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": "9c3b0db39639f529038d532ca1316cb7" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.5.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "1189ba33f356805ca674cc4d66d7b252" + }, + { + "dataPath": "params_shard_48.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": "7934666397ac8d94ffebadb5f7632627" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 28960768, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.6.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.7.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "13359b38b80084dfe3a94f165207d246" + }, + { + "dataPath": "params_shard_50.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": "194e6dd743ad70d3a9a2bf1c256dc28f" + }, + { + "dataPath": "params_shard_51.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": "13fe31e9bca0aa05a16d9c61f0dec445" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.8.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "00089bd24775ee8edb298fc07d60f729" + }, + { + "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": "ebda644edd7838884be52a2629153a42" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 28968960, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.9.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8139776 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8143872 + }, + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8147968 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19420160 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20829184 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23647232 + }, + { + "name": "model.layers.31.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23651328 + }, + { + "name": "model.layers.31.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23656448 + }, + { + "name": "model.layers.31.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26277888 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26605568 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28702720 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28964864 + } + ], + "md5sum": "b8fdc87fe3346ea1d173881d0028e1a8" + }, + { + "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": "8873af6074eb630a9911add44dd02674" + }, + { + "dataPath": "params_shard_56.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": "61ad27902bd88b5f0eb9ace6de4167da" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.32.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.32.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.32.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.32.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.33.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.33.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "2c1689c1472c171165bdc91a80927c45" + }, + { + "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": "f00b8595d196deee8535dff7f54a3e50" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 28960768, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.33.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.33.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.33.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.34.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.34.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.34.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.34.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.34.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "411020c5e6a50ccc7dc960c5b102a4cf" + }, + { + "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": "d21e8213c6018382d33455d1b57b8d7d" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 20820992, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.35.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.35.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.35.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.35.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.norm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + } + ], + "md5sum": "311b6fe905078f1b4aef390048a2c168" + } + ] +} \ No newline at end of file