| <?xml version="1.0"?>
|
| <net name="Model0" version="11">
|
| <layers>
|
| <layer id="0" name="image_feature" type="Parameter" version="opset1">
|
| <data shape="?,?,?" element_type="f32" />
|
| <output>
|
| <port id="0" precision="FP32" names="image_feature">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="1" name="pos_embed" type="Parameter" version="opset1">
|
| <data shape="?,?,?" element_type="f32" />
|
| <output>
|
| <port id="0" precision="FP32" names="pos_embed">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="2" name="key_padding_mask" type="Parameter" version="opset1">
|
| <data shape="?,?" element_type="f32" />
|
| <output>
|
| <port id="0" precision="FP32" names="key_padding_mask">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="3" name="Constant_9686" type="Const" version="opset1">
|
| <data element_type="i64" shape="4" offset="0" size="32" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>4</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="4" name="__module.attn/aten::view/Reshape_1" type="Reshape" version="opset1">
|
| <data special_zero="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>4</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="111">
|
| <dim>-1</dim>
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="5" name="__module.attn/aten::expand/Abs" type="Const" version="opset1">
|
| <data element_type="i64" shape="4" offset="32" size="32" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>4</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="6" name="__module.attn/aten::expand/Broadcast" type="Broadcast" version="opset3">
|
| <data mode="bidirectional" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>4</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="113">
|
| <dim>-1</dim>
|
| <dim>28</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="7" name="ShapeOf_9654" type="ShapeOf" version="opset3">
|
| <data output_type="i64" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="8" name="Constant_9655" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="64" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="9" name="Constant_9656" type="Const" version="opset1">
|
| <data element_type="i64" shape="" offset="64" size="8" />
|
| <output>
|
| <port id="0" precision="I64" />
|
| </output>
|
| </layer>
|
| <layer id="10" name="Gather_9657" type="Gather" version="opset8">
|
| <data batch_dims="0" />
|
| <input>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="2" precision="I64" />
|
| </input>
|
| <output>
|
| <port id="3" precision="I64" names="12,14,73">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="11" name="Constant_7566" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="72" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="12" name="__module.attn/aten::mul/Multiply" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="I64" names="104,105,114,115,92,93,98,99">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="13" name="Constant_326" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="80" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="14" name="Constant_9665" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="80" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="15" name="Constant_9666" type="Const" version="opset1">
|
| <data element_type="i64" shape="" offset="64" size="8" />
|
| <output>
|
| <port id="0" precision="I64" />
|
| </output>
|
| </layer>
|
| <layer id="16" name="Gather_9667" type="Gather" version="opset8">
|
| <data batch_dims="0" />
|
| <input>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="2" precision="I64" />
|
| </input>
|
| <output>
|
| <port id="3" precision="I64" names="103,109,97">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="17" name="__module.attn/prim::ListConstruct_2" type="Concat" version="opset1">
|
| <data axis="0" />
|
| <input>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="2" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="3" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="18" name="__module.attn/aten::reshape/Reshape" type="Reshape" version="opset1">
|
| <data special_zero="false" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>28</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="117,key_padding_mask_1">
|
| <dim>-1</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="19" name="aten::unsqueeze/Unsqueeze" type="Const" version="opset1">
|
| <data element_type="f32" shape="64, 1, 3584" offset="88" size="917504" />
|
| <output>
|
| <port id="0" precision="FP32" names="24">
|
| <dim>64</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="20" name="Constant_891" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="80" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="21" name="Constant_898" type="Const" version="opset1">
|
| <data element_type="i64" shape="1" offset="80" size="8" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="22" name="prim::ListConstruct/Concat" type="Concat" version="opset1">
|
| <data axis="0" />
|
| <input>
|
| <port id="0" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| <port id="2" precision="I64">
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="3" precision="I64" names="27">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="23" name="aten::repeat/Tile" type="Tile" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="28,query.1">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="24" name="__module.attn/prim::ListUnpack/VariadicSplit.0" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 3584" offset="917592" size="12845056" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="25" name="Convert_1993077" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="26" name="__module.attn/prim::ListUnpack/VariadicSplit.0/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="13762648" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="27" name="__module.attn/prim::ListUnpack/VariadicSplit.0/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="28" name="__module.attn/prim::ListUnpack/VariadicSplit.0/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="29" name="__module.attn/aten::linear/MatMul_1" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="30" name="Constant_9232" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="13769816" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="31" name="__module.attn/aten::linear/Add_1" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="89,q.3">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="32" name="Constant_9687" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="13784152" size="24" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="33" name="__module.attn/aten::view/Reshape_2" type="Reshape" version="opset1">
|
| <data special_zero="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="95">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="34" name="__module.attn/aten::transpose/Constant_1" type="Const" version="opset1">
|
| <data element_type="i32" shape="3" offset="13784176" size="12" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="35" name="__module.attn/aten::transpose/Transpose_1" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="96,q">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="36" name="Constant_9233" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 1" offset="13784188" size="4" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="37" name="__module.attn/aten::div/Divide_1" type="Divide" version="opset1">
|
| <data auto_broadcast="numpy" m_pythondiv="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="118,q_scaled">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="38" name="self.kv_proj.weight" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 1152" offset="13784192" size="4128768" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="39" name="Convert_2008158" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="40" name="self.kv_proj.weight/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="17912960" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="41" name="self.kv_proj.weight/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="42" name="self.kv_proj.weight/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="43" name="__module.kv_proj/ov_ext::linear/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>1152</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="41,input.1">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="44" name="__module.ln_kv/aten::layer_norm/Multiply" type="Const" version="opset1">
|
| <data element_type="i32" shape="1" offset="17920128" size="4" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="45" name="__module.ln_kv/aten::layer_norm/MVN" type="MVN" version="opset6">
|
| <data eps="9.9999999747524271e-07" normalize_variance="true" eps_mode="INSIDE_SQRT" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="46" name="Constant_9229" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="17920132" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="47" name="__module.ln_kv/aten::layer_norm/Multiply_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="48" name="Constant_9230" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="17934468" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="49" name="__module.ln_kv/aten::layer_norm/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="48">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="50" name="Constant_53" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="17948804" size="24" />
|
| <output>
|
| <port id="0" precision="I64" names="20">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="51" name="aten::permute/Transpose" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="21,image_feature_1">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="52" name="aten::add/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="30,key">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="53" name="__module.attn/prim::ListUnpack/VariadicSplit.1" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 3584" offset="17948828" size="12845056" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="54" name="Convert_1988050" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="55" name="__module.attn/prim::ListUnpack/VariadicSplit.1/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="30793884" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="56" name="__module.attn/prim::ListUnpack/VariadicSplit.1/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="57" name="__module.attn/prim::ListUnpack/VariadicSplit.1/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="58" name="__module.attn/aten::linear/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="59" name="Constant_9231" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="30801052" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="60" name="__module.attn/aten::linear/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="90,k.1">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="61" name="Constant_9688" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="30815388" size="24" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="62" name="__module.attn/aten::view/Reshape" type="Reshape" version="opset1">
|
| <data special_zero="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="101">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="63" name="__module.attn/aten::transpose/Constant" type="Const" version="opset1">
|
| <data element_type="i32" shape="3" offset="13784176" size="12" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="64" name="__module.attn/aten::transpose/Transpose" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="102,k">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="65" name="__module.attn/aten::baddbmm/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="66" name="__module.attn/aten::baddbmm/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>1</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="120,input.5">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="67" name="__module.attn/aten::softmax/Softmax" type="SoftMax" version="opset8">
|
| <data axis="-1" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32" names="121,attn_output_weights.1">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="68" name="__module.attn/prim::ListUnpack/VariadicSplit.2" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 3584" offset="30815412" size="12845056" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="69" name="Convert_1983023" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="70" name="__module.attn/prim::ListUnpack/VariadicSplit.2/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="43660468" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="71" name="__module.attn/prim::ListUnpack/VariadicSplit.2/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="72" name="__module.attn/prim::ListUnpack/VariadicSplit.2/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="73" name="__module.attn/aten::linear/MatMul_2" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="74" name="Constant_9234" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="43667636" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="75" name="__module.attn/aten::linear/Add_2" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="91,v.1">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="76" name="Constant_9689" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="30815388" size="24" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="77" name="__module.attn/aten::view/Reshape_3" type="Reshape" version="opset1">
|
| <data special_zero="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="107">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="78" name="__module.attn/aten::transpose/Constant_3" type="Const" version="opset1">
|
| <data element_type="i32" shape="3" offset="13784176" size="12" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="79" name="__module.attn/aten::transpose/Transpose_3" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="108,v">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="80" name="__module.attn/aten::bmm/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="false" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="122,attn_output.1">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="81" name="__module.attn/aten::transpose/Constant_4" type="Const" version="opset1">
|
| <data element_type="i32" shape="3" offset="13784176" size="12" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="82" name="__module.attn/aten::transpose/Transpose_4" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="123">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="83" name="Concat_4375" type="Const" version="opset1">
|
| <data element_type="i64" shape="2" offset="43681972" size="16" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>2</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="84" name="__module.attn/aten::view/Reshape_4" type="Reshape" version="opset1">
|
| <data special_zero="false" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>128</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>2</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="128">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="85" name="self.attn.out_proj.weight" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 3584" offset="43681988" size="12845056" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="86" name="Convert_1998104" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="87" name="self.attn.out_proj.weight/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="56527044" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="88" name="self.attn.out_proj.weight/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="89" name="self.attn.out_proj.weight/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="90" name="__module.attn.out_proj/ov_ext::linear/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="91" name="Constant_9235" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 3584" offset="56534212" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="92" name="__module.attn.out_proj/ov_ext::linear/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="131,attn_output">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="93" name="Constant_9690" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="56548548" size="24" />
|
| <output>
|
| <port id="0" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="94" name="__module.attn/aten::view/Reshape_5" type="Reshape" version="opset1">
|
| <data special_zero="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="134,out">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="95" name="Constant_486" type="Const" version="opset1">
|
| <data element_type="i64" shape="3" offset="17948804" size="24" />
|
| <output>
|
| <port id="0" precision="I64" names="35">
|
| <dim>3</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="96" name="aten::permute/Transpose_1" type="Transpose" version="opset1">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>64</dim>
|
| <dim>-1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I64">
|
| <dim>3</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="36,input">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="97" name="__module.ln_post/aten::layer_norm/Multiply" type="Const" version="opset1">
|
| <data element_type="i32" shape="1" offset="17920128" size="4" />
|
| <output>
|
| <port id="0" precision="I32">
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="98" name="__module.ln_post/aten::layer_norm/MVN" type="MVN" version="opset6">
|
| <data eps="9.9999999747524271e-07" normalize_variance="true" eps_mode="INSIDE_SQRT" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="I32">
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="99" name="Constant_9236" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="56548572" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="100" name="__module.ln_post/aten::layer_norm/Multiply_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="101" name="Constant_9237" type="Const" version="opset1">
|
| <data element_type="f32" shape="1, 1, 3584" offset="56562908" size="14336" />
|
| <output>
|
| <port id="0" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="102" name="__module.ln_post/aten::layer_norm/Add" type="Add" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>1</dim>
|
| <dim>1</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="141,x">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="103" name="Transpose_8047" type="Const" version="opset1">
|
| <data element_type="i8" shape="3584, 3584" offset="56577244" size="12845056" />
|
| <output>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="104" name="Convert_2003131" type="Convert" version="opset1">
|
| <data destination_type="f16" />
|
| <input>
|
| <port id="0" precision="I8">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="105" name="Transpose_8047/scale" type="Const" version="opset1">
|
| <data element_type="f16" shape="3584, 1" offset="69422300" size="7168" />
|
| <output>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="106" name="Transpose_8047/fq_weights_1" type="Multiply" version="opset1">
|
| <data auto_broadcast="numpy" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>1</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="107" name="Transpose_8047/fq_weights_1/convert" type="Convert" version="opset1">
|
| <data destination_type="f32" />
|
| <input>
|
| <port id="0" precision="FP16">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="108" name="aten::matmul/MatMul" type="MatMul" version="opset1">
|
| <data transpose_a="false" transpose_b="true" />
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| <port id="1" precision="FP32">
|
| <dim>3584</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| <output>
|
| <port id="2" precision="FP32" names="last_hidden_state">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </output>
|
| </layer>
|
| <layer id="109" name="Result_545" type="Result" version="opset1" output_names="last_hidden_state">
|
| <input>
|
| <port id="0" precision="FP32">
|
| <dim>-1</dim>
|
| <dim>64</dim>
|
| <dim>3584</dim>
|
| </port>
|
| </input>
|
| </layer>
|
| </layers>
|
| <edges>
|
| <edge from-layer="0" from-port="0" to-layer="7" to-port="0" />
|
| <edge from-layer="0" from-port="0" to-layer="43" to-port="0" />
|
| <edge from-layer="1" from-port="0" to-layer="52" to-port="1" />
|
| <edge from-layer="2" from-port="0" to-layer="4" to-port="0" />
|
| <edge from-layer="3" from-port="0" to-layer="4" to-port="1" />
|
| <edge from-layer="4" from-port="2" to-layer="6" to-port="0" />
|
| <edge from-layer="5" from-port="0" to-layer="6" to-port="1" />
|
| <edge from-layer="6" from-port="2" to-layer="18" to-port="0" />
|
| <edge from-layer="7" from-port="1" to-layer="10" to-port="0" />
|
| <edge from-layer="7" from-port="1" to-layer="16" to-port="0" />
|
| <edge from-layer="8" from-port="0" to-layer="10" to-port="1" />
|
| <edge from-layer="9" from-port="0" to-layer="10" to-port="2" />
|
| <edge from-layer="10" from-port="3" to-layer="12" to-port="0" />
|
| <edge from-layer="10" from-port="3" to-layer="22" to-port="1" />
|
| <edge from-layer="11" from-port="0" to-layer="12" to-port="1" />
|
| <edge from-layer="12" from-port="2" to-layer="17" to-port="0" />
|
| <edge from-layer="13" from-port="0" to-layer="17" to-port="1" />
|
| <edge from-layer="14" from-port="0" to-layer="16" to-port="1" />
|
| <edge from-layer="15" from-port="0" to-layer="16" to-port="2" />
|
| <edge from-layer="16" from-port="3" to-layer="17" to-port="2" />
|
| <edge from-layer="17" from-port="3" to-layer="18" to-port="1" />
|
| <edge from-layer="18" from-port="2" to-layer="66" to-port="0" />
|
| <edge from-layer="19" from-port="0" to-layer="23" to-port="0" />
|
| <edge from-layer="20" from-port="0" to-layer="22" to-port="0" />
|
| <edge from-layer="21" from-port="0" to-layer="22" to-port="2" />
|
| <edge from-layer="22" from-port="3" to-layer="23" to-port="1" />
|
| <edge from-layer="23" from-port="2" to-layer="29" to-port="0" />
|
| <edge from-layer="24" from-port="0" to-layer="25" to-port="0" />
|
| <edge from-layer="25" from-port="1" to-layer="27" to-port="0" />
|
| <edge from-layer="26" from-port="0" to-layer="27" to-port="1" />
|
| <edge from-layer="27" from-port="2" to-layer="28" to-port="0" />
|
| <edge from-layer="28" from-port="1" to-layer="29" to-port="1" />
|
| <edge from-layer="29" from-port="2" to-layer="31" to-port="0" />
|
| <edge from-layer="30" from-port="0" to-layer="31" to-port="1" />
|
| <edge from-layer="31" from-port="2" to-layer="33" to-port="0" />
|
| <edge from-layer="32" from-port="0" to-layer="33" to-port="1" />
|
| <edge from-layer="33" from-port="2" to-layer="35" to-port="0" />
|
| <edge from-layer="34" from-port="0" to-layer="35" to-port="1" />
|
| <edge from-layer="35" from-port="2" to-layer="37" to-port="0" />
|
| <edge from-layer="36" from-port="0" to-layer="37" to-port="1" />
|
| <edge from-layer="37" from-port="2" to-layer="65" to-port="0" />
|
| <edge from-layer="38" from-port="0" to-layer="39" to-port="0" />
|
| <edge from-layer="39" from-port="1" to-layer="41" to-port="0" />
|
| <edge from-layer="40" from-port="0" to-layer="41" to-port="1" />
|
| <edge from-layer="41" from-port="2" to-layer="42" to-port="0" />
|
| <edge from-layer="42" from-port="1" to-layer="43" to-port="1" />
|
| <edge from-layer="43" from-port="2" to-layer="45" to-port="0" />
|
| <edge from-layer="44" from-port="0" to-layer="45" to-port="1" />
|
| <edge from-layer="45" from-port="2" to-layer="47" to-port="0" />
|
| <edge from-layer="46" from-port="0" to-layer="47" to-port="1" />
|
| <edge from-layer="47" from-port="2" to-layer="49" to-port="0" />
|
| <edge from-layer="48" from-port="0" to-layer="49" to-port="1" />
|
| <edge from-layer="49" from-port="2" to-layer="51" to-port="0" />
|
| <edge from-layer="50" from-port="0" to-layer="51" to-port="1" />
|
| <edge from-layer="51" from-port="2" to-layer="52" to-port="0" />
|
| <edge from-layer="51" from-port="2" to-layer="73" to-port="0" />
|
| <edge from-layer="52" from-port="2" to-layer="58" to-port="0" />
|
| <edge from-layer="53" from-port="0" to-layer="54" to-port="0" />
|
| <edge from-layer="54" from-port="1" to-layer="56" to-port="0" />
|
| <edge from-layer="55" from-port="0" to-layer="56" to-port="1" />
|
| <edge from-layer="56" from-port="2" to-layer="57" to-port="0" />
|
| <edge from-layer="57" from-port="1" to-layer="58" to-port="1" />
|
| <edge from-layer="58" from-port="2" to-layer="60" to-port="0" />
|
| <edge from-layer="59" from-port="0" to-layer="60" to-port="1" />
|
| <edge from-layer="60" from-port="2" to-layer="62" to-port="0" />
|
| <edge from-layer="61" from-port="0" to-layer="62" to-port="1" />
|
| <edge from-layer="62" from-port="2" to-layer="64" to-port="0" />
|
| <edge from-layer="63" from-port="0" to-layer="64" to-port="1" />
|
| <edge from-layer="64" from-port="2" to-layer="65" to-port="1" />
|
| <edge from-layer="65" from-port="2" to-layer="66" to-port="1" />
|
| <edge from-layer="66" from-port="2" to-layer="67" to-port="0" />
|
| <edge from-layer="67" from-port="1" to-layer="80" to-port="0" />
|
| <edge from-layer="68" from-port="0" to-layer="69" to-port="0" />
|
| <edge from-layer="69" from-port="1" to-layer="71" to-port="0" />
|
| <edge from-layer="70" from-port="0" to-layer="71" to-port="1" />
|
| <edge from-layer="71" from-port="2" to-layer="72" to-port="0" />
|
| <edge from-layer="72" from-port="1" to-layer="73" to-port="1" />
|
| <edge from-layer="73" from-port="2" to-layer="75" to-port="0" />
|
| <edge from-layer="74" from-port="0" to-layer="75" to-port="1" />
|
| <edge from-layer="75" from-port="2" to-layer="77" to-port="0" />
|
| <edge from-layer="76" from-port="0" to-layer="77" to-port="1" />
|
| <edge from-layer="77" from-port="2" to-layer="79" to-port="0" />
|
| <edge from-layer="78" from-port="0" to-layer="79" to-port="1" />
|
| <edge from-layer="79" from-port="2" to-layer="80" to-port="1" />
|
| <edge from-layer="80" from-port="2" to-layer="82" to-port="0" />
|
| <edge from-layer="81" from-port="0" to-layer="82" to-port="1" />
|
| <edge from-layer="82" from-port="2" to-layer="84" to-port="0" />
|
| <edge from-layer="83" from-port="0" to-layer="84" to-port="1" />
|
| <edge from-layer="84" from-port="2" to-layer="90" to-port="0" />
|
| <edge from-layer="85" from-port="0" to-layer="86" to-port="0" />
|
| <edge from-layer="86" from-port="1" to-layer="88" to-port="0" />
|
| <edge from-layer="87" from-port="0" to-layer="88" to-port="1" />
|
| <edge from-layer="88" from-port="2" to-layer="89" to-port="0" />
|
| <edge from-layer="89" from-port="1" to-layer="90" to-port="1" />
|
| <edge from-layer="90" from-port="2" to-layer="92" to-port="0" />
|
| <edge from-layer="91" from-port="0" to-layer="92" to-port="1" />
|
| <edge from-layer="92" from-port="2" to-layer="94" to-port="0" />
|
| <edge from-layer="93" from-port="0" to-layer="94" to-port="1" />
|
| <edge from-layer="94" from-port="2" to-layer="96" to-port="0" />
|
| <edge from-layer="95" from-port="0" to-layer="96" to-port="1" />
|
| <edge from-layer="96" from-port="2" to-layer="98" to-port="0" />
|
| <edge from-layer="97" from-port="0" to-layer="98" to-port="1" />
|
| <edge from-layer="98" from-port="2" to-layer="100" to-port="0" />
|
| <edge from-layer="99" from-port="0" to-layer="100" to-port="1" />
|
| <edge from-layer="100" from-port="2" to-layer="102" to-port="0" />
|
| <edge from-layer="101" from-port="0" to-layer="102" to-port="1" />
|
| <edge from-layer="102" from-port="2" to-layer="108" to-port="0" />
|
| <edge from-layer="103" from-port="0" to-layer="104" to-port="0" />
|
| <edge from-layer="104" from-port="1" to-layer="106" to-port="0" />
|
| <edge from-layer="105" from-port="0" to-layer="106" to-port="1" />
|
| <edge from-layer="106" from-port="2" to-layer="107" to-port="0" />
|
| <edge from-layer="107" from-port="1" to-layer="108" to-port="1" />
|
| <edge from-layer="108" from-port="2" to-layer="109" to-port="0" />
|
| </edges>
|
| <rt_info>
|
| <Runtime_version value="2026.0.0-20965-c6d6a13a886-releases/2026/0" />
|
| <conversion_parameters>
|
| <framework value="pytorch" />
|
| <is_python_object value="True" />
|
| </conversion_parameters>
|
| <nncf>
|
| <friendly_names_were_updated value="True" />
|
| <version value="3.0.0" />
|
| <weight_compression>
|
| <advanced_parameters value="{'statistics_path': None, 'lora_adapter_rank': 256, 'group_size_fallback_mode': 'error', 'min_adjusted_group_size': 32, 'awq_params': {'subset_size': 32, 'percent_to_apply': 0.002, 'alpha_min': 0.0, 'alpha_max': 1.0, 'steps': 100, 'prefer_data_aware_scaling': True}, 'scale_estimation_params': {'subset_size': 64, 'initial_steps': 5, 'scale_steps': 5, 'weight_penalty': -1.0}, 'gptq_params': {'damp_percent': 0.1, 'block_size': 128, 'subset_size': 128}, 'lora_correction_params': {'adapter_rank': 8, 'num_iterations': 3, 'apply_regularization': True, 'subset_size': 128, 'use_int8_adapters': True}, 'backend_params': {}, 'codebook': None, 'adaptive_codebook_params': {'value_type': 'f8e4m3', 'across_blocks': False, 'num_elements': 16}}" />
|
| <all_layers value="False" />
|
| <awq value="False" />
|
| <backup_mode value="int8_asym" />
|
| <compression_format value="dequantize" />
|
| <gptq value="False" />
|
| <group_size value="-1" />
|
| <ignored_scope value="[]" />
|
| <lora_correction value="False" />
|
| <mode value="int8_sym" />
|
| <ratio value="1.0" />
|
| <scale_estimation value="False" />
|
| <sensitivity_metric value="weight_quantization_error" />
|
| </weight_compression>
|
| </nncf>
|
| <optimum>
|
| <nncf_version value="3.0.0" />
|
| <optimum_intel_version value="1.27.0" />
|
| <optimum_version value="2.1.0" />
|
| <pytorch_version value="2.11.0" />
|
| <transformers_version value="4.53.3" />
|
| </optimum>
|
| </rt_info>
|
| </net>
|
|
|
