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 +1,5857 @@ +{ + "metadata": { + "ParamSize": 518, + "ParamBytes": 1563781120.0, + "BitsPerParam": 4.506794107285474 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 64552960, + "records": [ + { + "name": "gpt_neox.embed_in.q_weight", + "shape": [ + 50432, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 64552960, + "byteOffset": 0 + } + ], + "md5sum": "1f177a96dad64173a6fabc6400aa2dc6" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 22855680, + "records": [ + { + "name": "gpt_neox.embed_in.q_scale", + "shape": [ + 50432, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8069120, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.0.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 8069120 + }, + { + "name": "gpt_neox.layers.0.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 8074240 + }, + { + "name": "gpt_neox.layers.0.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 8079360 + }, + { + "name": "gpt_neox.layers.0.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 8084480 + }, + { + "name": "gpt_neox.layers.0.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 8089600 + }, + { + "name": "gpt_neox.layers.0.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 17920000 + }, + { + "name": "gpt_neox.layers.0.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 19148800 + }, + { + "name": "gpt_neox.layers.0.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 19164160 + }, + { + "name": "gpt_neox.layers.0.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 22440960 + }, + { + "name": "gpt_neox.layers.0.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 22850560 + } + ], + "md5sum": "a6293b99693eabfe8f3ef4cd385d4e9e" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.0.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.0.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.0.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.0.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.0.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.0.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.1.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.1.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.1.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.1.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "74c91429fee09611d8b4b733735aaeae" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.1.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.1.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.1.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.1.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.1.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.1.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.1.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.1.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.1.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "eac5c034b57c479af8ae750eacb906bb" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.1.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.1.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.1.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.2.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.2.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.2.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.2.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.2.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.2.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.2.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.2.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.2.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.2.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "ad330c8fc91c3675e7f0ce8bc497ffee" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.2.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.2.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.2.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.2.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.2.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.2.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.3.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.3.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.3.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.3.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "1a1be58aa2b1a615f0a6e9091d6ee0cb" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.3.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.3.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.3.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.3.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.3.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.3.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.3.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.3.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.3.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "5e05efab1074b80a4a89e58db3296965" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.3.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.3.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.3.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.4.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.4.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.4.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.4.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.4.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.4.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.4.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.4.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.4.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.4.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "74362a3148d29fa83a8e4b651723a918" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.4.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.4.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.4.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.4.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.4.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.4.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.5.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.5.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.5.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.5.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "95b9047cfaea8c6f718e6f4dbc28d46f" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.5.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.5.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.5.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.5.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.5.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.5.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.5.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.5.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.5.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "fd51268c1c39de6f5d5dbe734cf6df0e" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.5.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.5.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.5.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.6.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.6.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.6.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.6.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.6.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.6.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.6.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.6.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.6.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.6.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "55c10e76bc7b42b2ce088309411011a7" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.6.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.6.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.6.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.6.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.6.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.6.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.7.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.7.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.7.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.7.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "eeb7f8581ddf8bcf6b1405688014f3d3" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.7.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.7.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.7.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.7.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.7.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.7.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.7.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.7.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.7.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "4bbb15807ca2c6a55d72bafc7c242e18" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.7.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.7.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.7.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.8.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.8.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.8.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.8.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.8.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.8.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.8.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.8.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.8.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.8.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "01b8bd794715cc42452619ab25330e61" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.8.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.8.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.8.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.8.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.8.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.8.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.9.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.9.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.9.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.9.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "3f8623ac0baf80b4b24aec5ff22b0c5b" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.9.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.9.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.9.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.9.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.9.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.9.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.9.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.9.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.9.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "058c4ac1cb0ef6c83e6fb41bc2884a8a" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.9.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.9.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.9.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.10.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.10.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.10.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.10.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.10.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.10.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.10.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.10.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.10.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.10.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "11b042257b002ebc2841ea642befbeff" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.10.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.10.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.10.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.10.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.10.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.10.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.11.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.11.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.11.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.11.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "5a15da59d06e50423f222619862c2856" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.11.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.11.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.11.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.11.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.11.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.11.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.11.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.11.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.11.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "20a2b4b5c089b0a3f03b98b0930bee6b" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.11.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.11.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.11.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.12.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.12.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.12.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.12.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.12.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.12.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.12.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.12.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.12.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.12.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "49e567c6132dcd6b242adb7ab89b21c3" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.12.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.12.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.12.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.12.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.12.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.12.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.13.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.13.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.13.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.13.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "82ce9ea7c7263bcfd6791b27034994f4" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.13.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.13.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.13.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.13.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.13.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.13.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.13.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.13.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.13.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "c7cdd5f24f13c456cc011f5af1d76bfa" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.13.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.13.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.13.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.14.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.14.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.14.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.14.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.14.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.14.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.14.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.14.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.14.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.14.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "7f92235ff7c16cbf5022aae65a622125" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.14.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.14.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.14.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.14.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.14.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.14.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.15.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.15.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.15.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.15.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "66724938fe90be6887a0952fa5d6ca20" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.15.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.15.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.15.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.15.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.15.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.15.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.15.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.15.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.15.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "7814c5fa9ba4fd2c3bb87968467cb97c" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.15.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.15.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.15.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.16.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.16.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.16.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.16.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.16.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.16.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.16.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.16.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.16.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.16.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "1f8c4edcf3c9f9b5c70a876c0882e553" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.16.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.16.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.16.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.16.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.16.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.16.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.17.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.17.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.17.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.17.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "38e60b4de635c31cd3d010da23de2e2b" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.17.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.17.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.17.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.17.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.17.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.17.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.17.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.17.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.17.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "0353b82959e9c1ff05dd3c382e204013" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.17.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.17.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.17.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.18.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.18.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.18.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.18.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.18.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.18.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.18.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.18.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.18.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.18.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "bd7665edd12c179ad173a4cb32ff0485" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.18.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.18.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.18.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.18.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.18.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.18.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.19.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.19.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.19.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.19.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "90d3fc82b19f4c94ef325ab344cb17e5" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.19.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.19.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.19.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.19.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.19.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.19.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.19.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.19.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.19.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "1e5405849d7df72035020a15293357f4" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.19.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.19.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.19.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.20.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.20.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.20.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.20.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.20.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.20.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.20.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.20.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.20.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.20.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "c53bc6b03ac6708cf015ccb371cb33b8" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.20.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.20.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.20.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.20.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.20.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.20.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.21.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.21.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.21.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.21.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "70bdbadf79c82ea466a9cfb0b7e4a776" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.21.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.21.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.21.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.21.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.21.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.21.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.21.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.21.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.21.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "5a2a7e6271febadbfe1478f445ef97c6" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.21.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.21.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.21.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.22.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.22.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.22.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.22.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.22.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.22.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.22.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.22.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.22.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.22.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "77725ed9a27568d47d63c2948c997c47" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.22.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.22.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.22.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.22.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.22.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.22.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.23.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.23.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.23.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.23.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "02d086a4ab1a1319d48cc6aa6f4c08e6" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.23.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.23.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.23.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.23.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.23.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.23.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.23.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.23.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.23.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "8dc8983eddd63bbd9f10ef2a9e05aa1f" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.23.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.23.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.23.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.24.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.24.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.24.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.24.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.24.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.24.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.24.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.24.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.24.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.24.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "32f187c72e77c00e9e15014802c0e466" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.24.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.24.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.24.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.24.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.24.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.24.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.25.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.25.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.25.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.25.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "7ea4a4512c6fa29a90fcafdad4cb9219" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.25.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.25.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.25.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.25.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.25.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.25.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.25.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.25.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.25.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "9a25e44bfa56fd0f3d08c0ea774b3407" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.25.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.25.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.25.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.26.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.26.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.26.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.26.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.26.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.26.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.26.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.26.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.26.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.26.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "53f0b67a3d89b16fe09b1683d76aa490" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.26.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.26.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.26.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.26.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.26.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.26.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.27.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.27.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.27.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.27.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "84d8019b28ac7d97ebded889c5e80478" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.27.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.27.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.27.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.27.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.27.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.27.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.27.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.27.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.27.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "6b3a8778276c1d11e3063e9a2f312ea1" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.27.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.27.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.27.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.28.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.28.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.28.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.28.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.28.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.28.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.28.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.28.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.28.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.28.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "5cc04808015bd999bbae65dc55a857ff" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.28.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.28.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.28.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.28.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.28.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.28.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.29.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.29.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.29.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.29.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "1275e3b3c73371139dd65a2b6931b3b6" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.29.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.29.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.29.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.29.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.29.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.29.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.29.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.29.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.29.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "99752a197f08eaed08e85ea577d03f07" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.29.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.29.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.29.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.30.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.layers.30.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "gpt_neox.layers.30.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.30.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.30.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 14771200 + }, + { + "name": "gpt_neox.layers.30.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 24601600 + }, + { + "name": "gpt_neox.layers.30.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 25830400 + }, + { + "name": "gpt_neox.layers.30.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 25845760 + }, + { + "name": "gpt_neox.layers.30.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 29122560 + }, + { + "name": "gpt_neox.layers.30.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "b430f919d2b41b5a87a45c56dc88b683" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 29537280, + "records": [ + { + "name": "gpt_neox.layers.30.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.30.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.30.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.layers.30.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.30.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.30.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 29511680 + }, + { + "name": "gpt_neox.layers.31.input_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29516800 + }, + { + "name": "gpt_neox.layers.31.input_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29521920 + }, + { + "name": "gpt_neox.layers.31.post_attention_layernorm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29527040 + }, + { + "name": "gpt_neox.layers.31.post_attention_layernorm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 29532160 + } + ], + "md5sum": "47174786839c4157185de62f04b89604" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 29532160, + "records": [ + { + "name": "gpt_neox.layers.31.attention.query_key_value.q_weight", + "shape": [ + 7680, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 9830400, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.31.attention.query_key_value.q_scale", + "shape": [ + 7680, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1228800, + "byteOffset": 9830400 + }, + { + "name": "gpt_neox.layers.31.attention.query_key_value.bias", + "shape": [ + 7680 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 15360, + "byteOffset": 11059200 + }, + { + "name": "gpt_neox.layers.31.attention.dense.q_weight", + "shape": [ + 2560, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 3276800, + "byteOffset": 11074560 + }, + { + "name": "gpt_neox.layers.31.attention.dense.q_scale", + "shape": [ + 2560, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 409600, + "byteOffset": 14351360 + }, + { + "name": "gpt_neox.layers.31.attention.dense.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14760960 + }, + { + "name": "gpt_neox.layers.31.mlp.dense_h_to_4h.q_weight", + "shape": [ + 10240, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 14766080 + }, + { + "name": "gpt_neox.layers.31.mlp.dense_h_to_4h.q_scale", + "shape": [ + 10240, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 27873280 + }, + { + "name": "gpt_neox.layers.31.mlp.dense_h_to_4h.bias", + "shape": [ + 10240 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 20480, + "byteOffset": 29511680 + } + ], + "md5sum": "f1dd2868c1a7b48c0d17dae9d0913414" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 64552960, + "records": [ + { + "name": "embed_out.q_weight", + "shape": [ + 50432, + 320 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 64552960, + "byteOffset": 0 + } + ], + "md5sum": "246047e05b693c90667bb4e3816fdf69" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 22830080, + "records": [ + { + "name": "gpt_neox.layers.31.mlp.dense_4h_to_h.q_weight", + "shape": [ + 2560, + 1280 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 13107200, + "byteOffset": 0 + }, + { + "name": "gpt_neox.layers.31.mlp.dense_4h_to_h.q_scale", + "shape": [ + 2560, + 320 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 1638400, + "byteOffset": 13107200 + }, + { + "name": "gpt_neox.layers.31.mlp.dense_4h_to_h.bias", + "shape": [ + 2560 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 5120, + "byteOffset": 14745600 + }, + { + "name": "gpt_neox.final_layer_norm.weight", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14750720 + }, + { + "name": "gpt_neox.final_layer_norm.bias", + "shape": [ + 2560 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 5120, + "byteOffset": 14755840 + }, + { + "name": "embed_out.q_scale", + "shape": [ + 50432, + 80 + ], + "dtype": "float16", + "format": "f32-to-bf16", + "nbytes": 8069120, + "byteOffset": 14760960 + } + ], + "md5sum": "5a59eb4bf9175cbd0a3462f3a82ab82a" + } + ] +} \ No newline at end of file