diff --git "a/ndarray-cache.json" "b/ndarray-cache.json" new file mode 100644--- /dev/null +++ "b/ndarray-cache.json" @@ -0,0 +1,4231 @@ +{ + "metadata": { + "ParamSize": 325, + "ParamBytes": 3457691648.0, + "BitsPerParam": 3.444661677836362 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 213417984, + "records": [ + { + "name": "lm_head.q_weight", + "shape": [ + 128256, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 213417984, + "byteOffset": 0 + } + ], + "md5sum": "0296bd52d43915b7ca3681be345179ae" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "1b5e9f167f99e2f911b9ab6ba9da60f0" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 213417984, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 128256, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 213417984, + "byteOffset": 0 + } + ], + "md5sum": "20e15bbf69af74ce784da70ea743f5cd" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "ee805a46a2f64ab065cf9cba582dae1f" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "dbde6834320fd2991bd27d918514e67b" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 32649216, + "records": [ + { + "name": "lm_head.q_scale", + "shape": [ + 128256, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 13338624, + "byteOffset": 0 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 13338624 + }, + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 13346816 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 14821376 + }, + { + "name": "model.norm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 14829568 + }, + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 128256, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 13338624, + "byteOffset": 14837760 + }, + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 28176384 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 28184576 + }, + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 29659136 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 32641024 + } + ], + "md5sum": "f20801273b5f8de8ecf2201086d78171" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "50cc13634697c2ad4c3d786e6b5c2ac4" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "15b441dbe3bd83f4d6e4c580188ccb03" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.0.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "42982c0014614a79ea4588fbda71e593" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "19f3c4f0040170d92403c5037927bc15" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "c32e266b6cb32338616f16db7daae714" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "e6acb63b1790bebde9b3c4e82efb7e58" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "4f60cca0a8f8382a10a41ed47eb50de5" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "88c471a5061280793ac0bfd9c902cecf" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "438195b5a19ffb2a0775ea5a6704058d" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "a5fcac6ef541d44e12e777d3dfe24807" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.3.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "65afc27ce7001ce0620d67f4f1425a15" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "d5e8a833a9ce027f921523ead41c6188" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "78c80861dde20f52fede95637bc5ca81" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "90678fe9e692e6acafc69a783d7aaf09" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "1087d9737f4a50afdfed75871847cae8" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "7734bc400f31f246a5be42670a5ff026" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "530424d3e296186b5e8b1fc2166066ca" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "df7cb9f4c3a78382de88c5af00a43e74" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.6.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "5cb7c9160716bbf3245f28c089daa9e1" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "29e50681bacc106362a74281843b15bb" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "d26f6459a76c47d46627c7cdfe9ff029" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "e83ba8adbf7caf3f9c613d9351e72e1a" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "deb5fd6b82d0bf7ea7e4b7d78a76ca0c" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "a1ed7c3eaba24c0ca234354247884b2c" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "17db9b95fe7f9e913d917d04ad9e8170" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "716c7756f0f8e821fdd8d53476f56bf1" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.10.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "1c7cba05e43d7894eb141a9af1d71349" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "df2806638c8d57cb7aeeb01f7fa52c73" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "5bcd04a7b239292d5d5c03bb397fb1e9" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "52bdab8b0d82675ca48e4a6b2639b339" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "940a1cb943d328f22da92843c88e2ce6" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "02067df98223335124ba6f033bc052d8" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "185cfa383031721981109a93d28e51ce" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "74b931313fcc0c9bb94bc9d61f56e9ac" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.13.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "c4b74f06d55350f74667301c22b72cc9" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "a4ad5b615b07aacdbf804159af4ff1ce" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "e8b209419e908996bf673f5c4a3e5f4f" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "0ff026d8831cfef851b5c74fae89af98" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "650b3981e2a142ba7f633682d830dfa6" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "fe7a00970c03690306503d2630894070" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "be81ccaf7a88ce73a5a7ac69234401fb" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "0ae6fd9bb90374522ec2aedd89e1b359" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.16.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "06994675c25ba4ab57627019f2328dab" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "24577cd6f7c6dd690195cb36138526d5" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "d58d9e3029f224ac51a2134f8b8231d0" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "c28d610f00dfa0300ffa4a97cfea3f98" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "001b7295e8fd6375e21c38ba3821182c" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "118bfdcb8a398e730e82d26081dd4d6a" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "5255507958121ec38ee42a903f88a93b" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 31948800, + "records": [ + { + "name": "model.layers.19.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 18104320 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 21086208 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 31309824 + } + ], + "md5sum": "9db445f0d0d84ffc39a467348e71ffd2" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "1114fefc088a02bd5a750b6e4ab73fa9" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "be4dfed8c732a9c7c992a0484880b775" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "3b726a5b8ec2880a3840151f368fdf32" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "a00a5f9508afe2aedc4c48097d246221" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "b64f64d37ff50c0729a2f7f3accfbd4c" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 27058176, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22577152 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22585344 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 22593536 + }, + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 24068096 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 27049984 + } + ], + "md5sum": "c651646c154ea5cadf9e2dd9e54bd5c5" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "434209bc7f1f6abce2ca87180cdb0eee" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "9f5be4223dcb895da08bbd98702a5532" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.21.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "9f3dd3e4a08b3c0e586aa074c197e5fd" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "6a4a81a913622a2dbd9e480b921b95a1" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "af9b1ee863e9320e5807698749aeb07d" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "6325a398d6972bab521d732ddb177ef1" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "74c4a6621f5fe64bddd79c3bdec496f1" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "e15a413918fa7188d0e2daf4963fba6a" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "3fa6c395425951eaa7428702894b965e" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "37b57d306da728cebf9e3ab27e95bc40" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.24.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "5ee12f57cc2907e04ffde2616f7dac5c" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "b0a4e4bda8d023c2e5a621c710b162d9" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "097d335ed895d868bd173e2c02d6fbdd" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "7f620339ef5e67cbc797d64700c985b8" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "6f51c156b700de8b36808e86a28f8934" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "bd9b455fbefbb44ed2be4e6c420d708a" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "d3a95526daf8f6127d6a88dee8643901" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "ec0f4f4f3207c7999b2a65e5150540e4" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 33439744, + "records": [ + { + "name": "model.layers.27.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 18104320 + }, + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 18112512 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 19587072 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 22568960 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 22577152 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 32800768 + } + ], + "md5sum": "63fbcd00d86c7f2e263049cc72d1a483" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "fa292ffa9462a72e7e31ececc75525f1" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 32317440, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 7241728 + }, + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 7249920 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 30842880 + } + ], + "md5sum": "4f552112ee502fb8ea966a9b6372cc25" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 23592960, + "records": [ + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 4096, + 1440 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 23592960, + "byteOffset": 0 + } + ], + "md5sum": "a9bc3f2fcbd8ac276aefc8e693a1a8cc" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "38b0b8366411f840af6fe8458a338281" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 25567232, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 0 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 2981888 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 2990080 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 13213696 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 13852672 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 20668416 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 21094400 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 4096, + 180 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1474560, + "byteOffset": 21102592 + }, + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 22577152 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 4096 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8192, + "byteOffset": 25559040 + } + ], + "md5sum": "12707481158fae38917613065a7ffa04" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 47710208, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 47710208, + "byteOffset": 0 + } + ], + "md5sum": "5b780cf61c71fd7a2f8b673ad09424c8" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 31948800, + "records": [ + { + "name": "model.layers.30.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 0 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 10223616 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 10862592 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 17678336 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 2981888, + "byteOffset": 18104320 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_weight", + "shape": [ + 6144, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 10223616, + "byteOffset": 21086208 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 6144, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 638976, + "byteOffset": 31309824 + } + ], + "md5sum": "56b0ff3b56354105401d196e6004ebde" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 7241728, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 4096, + 416 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 6815744, + "byteOffset": 0 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 4096, + 52 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 425984, + "byteOffset": 6815744 + } + ], + "md5sum": "03bbcb0a0fe7520999683024b1207363" + } + ] +} \ No newline at end of file