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": "6287072f47d913f5f880b82d6e647b4d" + }, + { + "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": "fb675d7dbe4875b2773aa04f5bc0c1d8" + }, + { + "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": "6b8207b8b9574817aadb033abd5ede70" + }, + { + "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": "5cfdf5fb5de126a42d7b51e7c034c0e7" + }, + { + "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": "c043f717a9640198fa11176e4418af22" + }, + { + "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": "ca28406ebc06549e494b4194c913bc46" + }, + { + "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": "263befdfed5997220a3bd227f100c17b" + }, + { + "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": "7cc9e5809b1f6fba38f8f71eca66d3bc" + }, + { + "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": "6f5d92b268a1d86589e8339511d7876a" + }, + { + "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": "b2c27b16d2e4bd32922644f431ce1174" + }, + { + "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": "6b064686bd0cd78f8f0fb9358927835f" + }, + { + "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": "a7ef22f86c7e2c22d2e599c958a11133" + }, + { + "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": "32989531738bc29cef7a4d3d067f3d42" + }, + { + "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": "598c0a2a6f3a4f615ef6a79c2ea1dfda" + }, + { + "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": "f700016af23cdc2f0a10eb739469876b" + }, + { + "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": "a6618be2d757b28725424673cc631aba" + }, + { + "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": "a744507c089792b723d138e2aa2e1897" + }, + { + "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": "c2bb9a35a443867ac976b2828f68b243" + }, + { + "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": "e4a367875748f03279171ae63625a0a2" + }, + { + "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": "406b5945c8949582b35a4789c28f882f" + }, + { + "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": "567c15dc54fc28422991fd6eadf990b2" + }, + { + "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": "bcaaadcf15c0be4665af2fb6db181d26" + }, + { + "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": "074bb862e9163f216d2199a1bd22dc79" + }, + { + "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": "ceb7a9aba617b31304387b03d9ff31bf" + }, + { + "dataPath": "params_shard_24.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": "c22e6afe51b3401cb0fafb0b1086abcd" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 31788032, + "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 + }, + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 28960768 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 31778816 + }, + { + "name": "model.layers.21.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 31782912 + } + ], + "md5sum": "ef25a08df9ba226ec28b2afdf36d63f9" + }, + { + "dataPath": "params_shard_26.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": "df0f3841409baaf8d1fe84d0a75cf063" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 26133504, + "records": [ + { + "name": "model.layers.21.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 0 + }, + { + "name": "model.layers.21.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 2621440 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 2949120 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 5046272 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 5308416 + }, + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 5312512 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 16584704 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 17993728 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20811776 + }, + { + "name": "model.layers.3.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 20815872 + }, + { + "name": "model.layers.3.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 20820992 + }, + { + "name": "model.layers.3.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 23442432 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 23770112 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 25867264 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 26129408 + } + ], + "md5sum": "10a4cf7890baee613eac0aab7e503f8f" + }, + { + "dataPath": "params_shard_28.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": "315fe78ac5d9908ef92188f41c40b475" + }, + { + "dataPath": "params_shard_29.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": "4bf9fc67a42e73fc9df9568e0c80f7f9" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.4.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.4.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "715e236badb96d306870dd23f47fa0d9" + }, + { + "dataPath": "params_shard_31.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": "65869f0cc3215f8114aaf4252fdaebd0" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 28960768, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.5.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.5.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.6.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.6.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "d784611f6f0665ab76eb7fcd88674c04" + }, + { + "dataPath": "params_shard_33.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": "820a8cc71d81e0e458704a74569c76d5" + }, + { + "dataPath": "params_shard_34.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": "23c9da940e6e3e04dfa36a4829a56ac9" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.7.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.7.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "5eb40c6adda54909244a8fb3c0c33d69" + }, + { + "dataPath": "params_shard_36.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": "e241e7cf9b0e9de3c0368f573c2558f1" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 28956672, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.8.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.8.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.9.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.9.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + } + ], + "md5sum": "65df2536a7035db83319f0495a8eefe1" + }, + { + "dataPath": "params_shard_38.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": "e4865c7d52a9c0a0ab35af9073d22446" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 33506304, + "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.22.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 12681216 + }, + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 12685312 + }, + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 23957504 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 25366528 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28184576 + }, + { + "name": "model.layers.22.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 28188672 + }, + { + "name": "model.layers.22.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 28193792 + }, + { + "name": "model.layers.22.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 30815232 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 31142912 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 33240064 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 33502208 + } + ], + "md5sum": "463a4d6f186e9c86770ccf300043f46e" + }, + { + "dataPath": "params_shard_40.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": "aacea256c60cd2c6985317af9c40cf83" + }, + { + "dataPath": "params_shard_41.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": "6d0cc0c164a5fc55923c523135bf4cd1" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.23.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.23.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.23.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "2acf0186a4ff2ceed08aad9830d7d8cd" + }, + { + "dataPath": "params_shard_43.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": "8dbbcd1433c7007cc0a2ceef208c7d11" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 28960768, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.24.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.24.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.24.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.25.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.25.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.25.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "f14ea30e6d3306de22ed2ef8cf6b1a35" + }, + { + "dataPath": "params_shard_45.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": "f3df185a7a5724bfd1bbac17ed5f9302" + }, + { + "dataPath": "params_shard_46.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": "ce2da723ccfb95c8608df3ded077c571" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.26.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.26.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.26.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "227e7f3f80fda9b9239a7e5b35919057" + }, + { + "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": "c4afa89e6f8ef18465e96ee987bd49c3" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 28960768, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.27.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.27.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.27.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.28.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.28.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.28.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "3b8aa76bbcaec4af2dd5ca222ff63829" + }, + { + "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": "54df372917b0015a711b27b5c08b846f" + }, + { + "dataPath": "params_shard_51.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": "63f7390766dbde60ce1d5c4c2411147d" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 33502208, + "records": [ + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 0 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.29.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 15503360 + }, + { + "name": "model.layers.29.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 15508480 + }, + { + "name": "model.layers.29.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 18129920 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 18457600 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 20816896 + }, + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 20820992 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "e2cb5fa23b8953a66d367ddf5ba65ef6" + }, + { + "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": "080b42342fcdbab33c374a1bec585ffb" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 28960768, + "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.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.30.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 2822144 + }, + { + "name": "model.layers.30.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 2827264 + }, + { + "name": "model.layers.30.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 5448704 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 5776384 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 8135680 + }, + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 2048, + 1376 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 11272192, + "byteOffset": 8139776 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 2048, + 344 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.31.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 23643136 + }, + { + "name": "model.layers.31.self_attn.c_attn.q_weight", + "shape": [ + 2560, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2621440, + "byteOffset": 23648256 + }, + { + "name": "model.layers.31.self_attn.c_attn.q_scale", + "shape": [ + 2560, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 327680, + "byteOffset": 26269696 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 2048, + 256 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 2097152, + "byteOffset": 26597376 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 2048, + 64 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "4f178cce800d5de72c39ae64259f5093" + }, + { + "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": "068ccb6913af7e7a040b531b443fcb4e" + }, + { + "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": "6a1df095ceefb8d6d42fa7e7b7ba277b" + }, + { + "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": "c592692a9878274f5523920c0eb3ef6d" + }, + { + "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": "81389fee68af06ff8947113fdb3d5596" + }, + { + "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": "a5afa80516a6a41bdae3f8c1be4bfd8e" + }, + { + "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": "c8521d65c09311be7faac30a794dfee3" + }, + { + "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": "32c3382d7c4dd530fcdd6d3abd54ac66" + } + ] +} \ No newline at end of file