diff --git "a/ndarray-cache-b16.json" "b/ndarray-cache-b16.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache-b16.json" @@ -0,0 +1,4785 @@ +{ + "metadata": { + "ParamSize": 399, + "ParamBytes": 1929527296.0, + "BitsPerParam": 5.002114406234114 + }, + "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": "b4b3aa0293e794582df5200974950158" + }, + { + "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": "fa76bbfdc3c178310719a0adbe7d712c" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 32133120, + "records": [ + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 151936, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 19447808, + "byteOffset": 0 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 30724096 + } + ], + "md5sum": "b93ff66a74349e2aa7d3c6c582fa0082" + }, + { + "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": "3fd0e3d48dc63f8616db74c34e523f71" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.0.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.1.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "6fd513c63709b6a6d881505be0db5857" + }, + { + "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": "12960fe711cd59cbc34fb35beb389f49" + }, + { + "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": "e640a3dbd801f37ff97429e160b409f0" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.10.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "3ba0f2b8385b194e38cb9ecb6823dc46" + }, + { + "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": "12ab76c93850a2592a3e42a81be49548" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.11.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.12.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "8165ba22d6f916cc5abf99b355a2b949" + }, + { + "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": "26479fb547b75e7b3ae270915f46a4d2" + }, + { + "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": "dbdf493d207230d050f7294b0e78ffe0" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.13.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "154494603f6986761abeaa10285e7510" + }, + { + "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": "ebb05bc7592b1d9460f43eb7b28c5c72" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.14.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.15.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "4069926e311684c765b3ba8687106f35" + }, + { + "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": "f19f079ea61bc4a5b76d57edc5b32223" + }, + { + "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": "0b478b5c45538a2f8b3f1f81a6067b37" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.16.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "af58419c71c67134f34a8a7e063c699f" + }, + { + "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": "0ffdf1a4cdafb0d3dc5a5aa79d67d298" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.17.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.18.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "7714565003127122d4f04169216e02b0" + }, + { + "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": "ef91e9b0ee3d8b6d1c1377fd7dcb012f" + }, + { + "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": "cd5ed55a75f66bb0f3e9f7667b2863d8" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.19.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "6365a59bf9c4d08e34ddff78b3a541cf" + }, + { + "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": "4546a15cccc08a393062b982c5d0a3ea" + }, + { + "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": "43dd9957f642388a1187c7d65a41820f" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.2.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.20.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + }, + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 28960768 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 31778816 + }, + { + "name": "model.layers.21.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 31782912 + } + ], + "md5sum": "8c6e80a34732bb9665f8e4f0dbd36f4a" + }, + { + "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": "33ad7e836f97ef66864cf39681e691b9" + }, + { + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 5046272 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 16584704 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 17993728 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 20811776 + }, + { + "name": "model.layers.3.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 25867264 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 26129408 + } + ], + "md5sum": "32cce9296abbc2072dbea3e89c6d27e8" + }, + { + "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": "2233f4b27adfbd85ba9a6889e3df6b3a" + }, + { + "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": "c8deeec30a491a1974d921ead94ffa33" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.4.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "976add38ab8d66e39a8fe372af915624" + }, + { + "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": "7baba4c3a77f18a8188aa2c59178a275" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.5.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.6.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "6bfeeded8e5b45fa8ea8036e08516ab4" + }, + { + "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": "d01035db81d0f2ecf64e7712df49878f" + }, + { + "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": "0e0f383473afee2714c0b1924b42f15a" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.7.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "e461d8ad8b0980827ac03342583f05aa" + }, + { + "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": "be40cf6621591d9ef636044fff2e37ac" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.8.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.9.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + } + ], + "md5sum": "92f809f83456238626ab72baa9c09840" + }, + { + "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": "15c85ef155bae410e7472dfdf715529b" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 23957504 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 25366528 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28184576 + }, + { + "name": "model.layers.22.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 33240064 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 33502208 + } + ], + "md5sum": "8eda1779b9d6168449bbaa6018d13bef" + }, + { + "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": "6265155b4713afdae15a55fc15212760" + }, + { + "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": "3137ed94263f77d091431e7e82a2796b" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.23.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "037cbef20d8399e8f2347b0758a501ef" + }, + { + "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": "eeab6499699e83339552860775c708ce" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.24.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.25.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "5bb5b70b8e15326ec4f2bcdfaf9b8618" + }, + { + "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": "8319a28142bfe6b25135a13ef57b543d" + }, + { + "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": "fc4abbc007e7f181c26f91b1a3163fc6" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.26.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "e73a1de70ce2052097cc557728a5f921" + }, + { + "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": "1c3006c321518f57263615d0f3fc11da" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.27.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.28.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "4e45649b86b28d1558909e3f73da5d0d" + }, + { + "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": "b647975f864d2d34a06f02ab4d23b46e" + }, + { + "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": "d9a31620f446d163ce25b7c5714c3d6b" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.29.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "e6afc1be622c8839571363afb7b7714f" + }, + { + "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": "9d4a3ea8b879b8ad5bef1657443d9155" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.30.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.31.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "fb7e36839fd2f3b20dd50efefeed8964" + }, + { + "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": "76a258c55e66351ce3aad4e8afba07de" + }, + { + "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": "1dbfca95c8ad943c5a0a8751115cbad6" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.32.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 32093184 + } + ], + "md5sum": "044bd81476ccf6dc4bf8474866d213b4" + }, + { + "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": "eddc4a46d00be78b0f3e8da3877f98e0" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 0 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 2818048 + }, + { + "name": "model.layers.33.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 7873536 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 19411968 + }, + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 20820992 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 23639040 + }, + { + "name": "model.layers.34.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 28694528 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 28956672 + } + ], + "md5sum": "0a4717da3e1ec5d51ed02e2dd53dbc5c" + }, + { + "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": "2382cbc1abeaa47eef84eec01930e402" + }, + { + "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": "bfloat16", + "format": "raw", + "nbytes": 1409024, + "byteOffset": 11272192 + }, + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 22016, + 64 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 2818048, + "byteOffset": 12681216 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 15499264 + }, + { + "name": "model.layers.35.self_attn.c_attn.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "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": "bfloat16", + "format": "raw", + "nbytes": 262144, + "byteOffset": 20554752 + }, + { + "name": "model.norm.weight", + "shape": [ + 2048 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 4096, + "byteOffset": 20816896 + } + ], + "md5sum": "36ad9c4159afe5eb503d1fe612678da6" + } + ] +} \ No newline at end of file