tiny-random-CohereForCausalLM / openvino_model.xml
h0witended's picture
Add reference OpenVINO IRs
a621156
Raw History Blame
150 kB
<?xml version="1.0"?>
<net name="Model0" version="11">
<layers>
<layer id="0" name="input_ids" type="Parameter" version="opset1">
<data shape="?,?" element_type="i64" />
<output>
<port id="0" precision="I64" names="input_ids">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="1" name="attention_mask" type="Parameter" version="opset1">
<data shape="?,?" element_type="i64" />
<output>
<port id="0" precision="I64" names="attention_mask">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="2" name="position_ids" type="Parameter" version="opset1">
<data shape="?,?" element_type="i64" />
<output>
<port id="0" precision="I64" names="position_ids">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="3" name="beam_idx" type="Parameter" version="opset1">
<data shape="?" element_type="i32" />
<output>
<port id="0" precision="I32" names="beam_idx">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="4" name="Constant_12272" type="Const" version="opset1">
<data element_type="f32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="FP32" />
</output>
</layer>
<layer id="5" name="ShapeOf_12258" type="ShapeOf" version="opset3">
<data output_type="i64" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="I64">
<dim>2</dim>
</port>
</output>
</layer>
<layer id="6" name="Constant_12260" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="7" name="Constant_12262" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="8" name="Gather_12263" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I64">
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="9" name="Constant_12265" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="16" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="10" name="Constant_12267" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="11" name="Constant_12269" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="24" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="12" name="Concat_12270" 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>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="4" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="13" name="Broadcast_12273" type="Broadcast" version="opset3">
<data mode="numpy" />
<input>
<port id="0" precision="FP32" />
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="14" name="ReadValue_12089" type="ReadValue" version="opset6">
<data variable_id="past_key_values.1.valuepresent.1.value" variable_type="f32" variable_shape="?,4,?,8" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="past_key_values.1.value">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="15" name="Constant_11914" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="16" name="Gather_11915" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="17" name="self.model.embed_tokens.weight" type="Const" version="opset1">
<data element_type="f32" shape="1024, 32" offset="32" size="131072" />
<output>
<port id="0" precision="FP32" names="self.model.embed_tokens.weight">
<dim>1024</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="18" name="__module.model.embed_tokens/aten::embedding/Convert" type="Convert" version="opset1">
<data destination_type="i32" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="19" name="__module.model.embed_tokens/aten::embedding/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="20" name="__module.model.embed_tokens/aten::embedding/Gather" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>1024</dim>
<dim>32</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="FP32" names="157,75,hidden_states.1,inputs_embeds">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="21" name="Constant_353" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="131104" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="22" name="__module.model.layers.0.input_layernorm/aten::mul/Multiply" type="MVN" version="opset6">
<data eps="9.9999997473787516e-06" normalize_variance="true" eps_mode="INSIDE_SQRT" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="167,169,170,hidden_states.3,hidden_states.5,hidden_states.7">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="23" name="self.model.layers.0.self_attn.q_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="131112" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.self_attn.q_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="24" name="__module.model.layers.0.self_attn.q_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="181">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="25" name="Constant_10863" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="26" name="__module.model.layers.0.self_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>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="183,query_states.1">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="27" name="__module.model.layers.0.self_attn/aten::transpose/Constant" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="28" name="__module.model.layers.0.self_attn/aten::to/Convert" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="195,q.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="29" name="__module.model.rotary_emb/aten::to/Convert" type="Const" version="opset1">
<data element_type="f32" shape="1, 4, 1" offset="135256" size="16" />
<output>
<port id="0" precision="FP32" names="131">
<dim>1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="30" name="61" type="Const" version="opset1">
<data element_type="i64" shape="" offset="135272" size="8" />
<output>
<port id="0" precision="I64" names="61" />
</output>
</layer>
<layer id="31" name="__module.model.rotary_emb/aten::unsqueeze/Unsqueeze_2" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="136,137">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="32" name="__module.model.rotary_emb/aten::to/Convert_2" type="Convert" version="opset1">
<data destination_type="f32" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="138,140,position_ids_expanded,position_ids_expanded.1">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="33" name="__module.model.rotary_emb/aten::matmul/MatMul" type="MatMul" version="opset1">
<data transpose_a="false" transpose_b="false" />
<input>
<port id="0" precision="FP32">
<dim>1</dim>
<dim>4</dim>
<dim>1</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="141">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="34" name="__module.model.rotary_emb/aten::transpose/Constant" type="Const" version="opset1">
<data element_type="i32" shape="3" offset="135280" size="12" />
<output>
<port id="0" precision="I32">
<dim>3</dim>
</port>
</output>
</layer>
<layer id="35" name="__module.model.rotary_emb/aten::transpose/Transpose" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I32">
<dim>3</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="142,freqs">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="36" name="__module.model.rotary_emb/aten::repeat_interleave/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="37" name="__module.model.rotary_emb/aten::repeat_interleave/ShapeOf" type="ShapeOf" version="opset3">
<data output_type="i32" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>3</dim>
</port>
</output>
</layer>
<layer id="38" name="55" type="Const" version="opset1">
<data element_type="i64" shape="" offset="131104" size="8" />
<output>
<port id="0" precision="I64" names="55" />
</output>
</layer>
<layer id="39" name="__module.model.rotary_emb/aten::repeat_interleave/Gather" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I32">
<dim>3</dim>
</port>
<port id="1" precision="I64" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I32" />
</output>
</layer>
<layer id="40" name="__module.model.rotary_emb/aten::repeat_interleave/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135292" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="41" name="__module.model.rotary_emb/aten::repeat_interleave/Range" type="Range" version="opset4">
<data output_type="i32" />
<input>
<port id="0" precision="I32" />
<port id="1" precision="I32" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="42" name="__module.model.rotary_emb/aten::repeat_interleave/Unsqueeze" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I32">
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="I32">
<dim>1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="43" name="__module.model.rotary_emb/aten::repeat_interleave/Concat" type="Const" version="opset1">
<data element_type="i32" shape="2" offset="135296" size="8" />
<output>
<port id="0" precision="I32">
<dim>2</dim>
</port>
</output>
</layer>
<layer id="44" name="__module.model.rotary_emb/aten::repeat_interleave/Tile" type="Tile" version="opset1">
<input>
<port id="0" precision="I32">
<dim>1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32">
<dim>2</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>2</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="45" name="__module.model.rotary_emb/aten::repeat_interleave/Constant_4" type="Const" version="opset1">
<data element_type="i32" shape="2" offset="135272" size="8" />
<output>
<port id="0" precision="I32">
<dim>2</dim>
</port>
</output>
</layer>
<layer id="46" name="__module.model.rotary_emb/aten::repeat_interleave/Transpose" type="Transpose" version="opset1">
<input>
<port id="0" precision="I32">
<dim>2</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32">
<dim>2</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>4</dim>
<dim>2</dim>
</port>
</output>
</layer>
<layer id="47" name="Constant_5032" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="131104" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="48" name="__module.model.rotary_emb/aten::repeat_interleave/Reshape_1" type="Reshape" version="opset1">
<data special_zero="false" />
<input>
<port id="0" precision="I32">
<dim>4</dim>
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>8</dim>
</port>
</output>
</layer>
<layer id="49" name="__module.model.rotary_emb/aten::repeat_interleave/Gather_1" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32">
<dim>8</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="FP32" names="143,emb">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="50" name="__module.model.rotary_emb/aten::cos/Cos" type="Cos" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="144,145,148,cos.1,cos.3">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="51" name="__module.model.layers.0.self_attn/aten::unsqueeze/Unsqueeze" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="FP32" names="197,296,cos,cos.5">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="52" name="__module.model.layers.0.self_attn/aten::mul/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="199">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="53" name="Constant_585" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="54" name="Constant_587" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="55" name="Constant_589" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="56" name="Constant_583" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="57" name="__module.model.layers.0.self_attn/aten::slice/Slice" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="201,x2.1">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="58" name="__module.model.layers.0.self_attn/aten::neg/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="59" name="__module.model.layers.0.self_attn/aten::neg/ConvertLike" type="Convert" version="opset1">
<data destination_type="f32" />
<input>
<port id="0" precision="I32" />
</input>
<output>
<port id="1" precision="FP32" />
</output>
</layer>
<layer id="60" name="__module.model.layers.0.self_attn/aten::neg/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="FP32" />
</input>
<output>
<port id="2" precision="FP32" names="202">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="61" name="__module.model.layers.0.self_attn/aten::stack/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="62" name="__module.model.layers.0.self_attn/aten::stack/Unsqueeze" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="63" name="Constant_575" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="64" name="Constant_577" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="65" name="Constant_579" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="66" name="Constant_573" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="67" name="__module.model.layers.0.self_attn/aten::slice/Slice_1" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="200,x1.1">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="68" name="__module.model.layers.0.self_attn/aten::stack/Unsqueeze_1" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="69" name="__module.model.layers.0.self_attn/aten::stack/Concat" type="Concat" version="opset1">
<data axis="-1" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="204">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
</output>
</layer>
<layer id="70" name="Constant_10865" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135336" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="71" name="__module.model.layers.0.self_attn/aten::flatten/Reshape" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="205">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="72" name="__module.model.rotary_emb/aten::sin/Sin" type="Sin" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="146,147,149,sin.1,sin.3">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="73" name="__module.model.layers.0.self_attn/aten::unsqueeze/Unsqueeze_1" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="FP32" names="198,297,sin,sin.5">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="74" name="__module.model.layers.0.self_attn/aten::mul/Multiply_1" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="206">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="75" name="__module.model.layers.0.self_attn/aten::add/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="207,217,q_embed.1,query.1">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="76" name="Constant_12284" type="Const" version="opset1">
<data element_type="f32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="FP32" />
</output>
</layer>
<layer id="77" name="Constant_12277" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="16" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="78" name="Constant_12279" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="79" name="Constant_12281" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="24" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="80" name="Concat_12282" 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>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="4" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="81" name="Broadcast_12285" type="Broadcast" version="opset3">
<data mode="numpy" />
<input>
<port id="0" precision="FP32" />
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="82" name="ReadValue_12083" type="ReadValue" version="opset6">
<data variable_id="past_key_values.0.keypresent.0.key" variable_type="f32" variable_shape="?,4,?,8" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="past_key_values.0.key">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="83" name="Constant_11905" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="84" name="Gather_11906" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="85" name="self.model.layers.0.self_attn.k_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="135368" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.self_attn.k_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="86" name="__module.model.layers.0.self_attn.k_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="185">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="87" name="Constant_10864" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="88" name="__module.model.layers.0.self_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>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="187,key_states.1">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="89" name="__module.model.layers.0.self_attn/aten::transpose/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="90" name="__module.model.layers.0.self_attn/aten::to/Convert_2" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="196,k.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="91" name="__module.model.layers.0.self_attn/aten::mul/Multiply_2" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="208">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="92" name="Constant_639" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="93" name="Constant_641" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="94" name="Constant_643" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="95" name="Constant_637" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="96" name="__module.model.layers.0.self_attn/aten::slice/Slice_2" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="210,x2.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="97" name="__module.model.layers.0.self_attn/aten::neg/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="98" name="__module.model.layers.0.self_attn/aten::neg/ConvertLike_1" type="Convert" version="opset1">
<data destination_type="f32" />
<input>
<port id="0" precision="I32" />
</input>
<output>
<port id="1" precision="FP32" />
</output>
</layer>
<layer id="99" name="__module.model.layers.0.self_attn/aten::neg/Multiply_1" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="FP32" />
</input>
<output>
<port id="2" precision="FP32" names="211">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="100" name="__module.model.layers.0.self_attn/aten::stack/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="101" name="__module.model.layers.0.self_attn/aten::stack/Unsqueeze_2" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="102" name="Constant_629" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="103" name="Constant_631" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="104" name="Constant_633" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="105" name="Constant_627" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="106" name="__module.model.layers.0.self_attn/aten::slice/Slice_3" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="209,x1.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="107" name="__module.model.layers.0.self_attn/aten::stack/Unsqueeze_3" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="108" name="__module.model.layers.0.self_attn/aten::stack/Concat_1" type="Concat" version="opset1">
<data axis="-1" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="213">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
</output>
</layer>
<layer id="109" name="Constant_10866" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135336" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="110" name="__module.model.layers.0.self_attn/aten::flatten/Reshape_1" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="214">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="111" name="__module.model.layers.0.self_attn/aten::mul/Multiply_3" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="215">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="112" name="__module.model.layers.0.self_attn/aten::add/Add_1" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="216,218,k_embed.1,key_states.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="113" name="__module.model.layers.0.self_attn/aten::cat/Concat" type="Concat" version="opset1">
<data axis="-2" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="hidden_states.9">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="114" name="Constant_12296" type="Const" version="opset1">
<data element_type="f32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="FP32" />
</output>
</layer>
<layer id="115" name="Constant_12289" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="16" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="116" name="Constant_12291" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="117" name="Constant_12293" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="24" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="118" name="Concat_12294" 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>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="4" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="119" name="Broadcast_12297" type="Broadcast" version="opset3">
<data mode="numpy" />
<input>
<port id="0" precision="FP32" />
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="120" name="ReadValue_12085" type="ReadValue" version="opset6">
<data variable_id="past_key_values.0.valuepresent.0.value" variable_type="f32" variable_shape="?,4,?,8" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="past_key_values.0.value">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="121" name="Constant_11908" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="122" name="Gather_11909" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="123" name="self.model.layers.0.self_attn.v_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="139464" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.self_attn.v_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="124" name="__module.model.layers.0.self_attn.v_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="189">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="125" name="Constant_10867" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="126" name="__module.model.layers.0.self_attn/aten::view/Reshape_2" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="191,value_states.1">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="127" name="__module.model.layers.0.self_attn/aten::transpose/Constant_2" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="128" name="__module.model.layers.0.self_attn/aten::transpose/Transpose_2" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="194,value_states.3">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="129" name="__module.model.layers.0.self_attn/aten::cat/Concat_1" type="Concat" version="opset1">
<data axis="-2" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="hidden_states.11">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="130" name="__module.model/aten::new_ones/Broadcast" type="Const" version="opset1">
<data element_type="boolean" shape="" offset="143560" size="1" />
<output>
<port id="0" precision="BOOL" names="112,result.1" />
</output>
</layer>
<layer id="131" name="__module.model/aten::arange/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="132" name="ShapeOf_10803" type="ShapeOf" version="opset3">
<data output_type="i64" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="133" name="Constant_10804" type="Const" version="opset1">
<data element_type="i64" shape="" offset="135312" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="134" name="Constant_10805" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="135" name="Gather_10806" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I64">
<dim>4</dim>
</port>
<port id="1" precision="I64" />
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="I64" names="79,81,83,q_offset.3" />
</output>
</layer>
<layer id="136" name="ShapeOf_10811" type="ShapeOf" version="opset3">
<data output_type="i64" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="1" precision="I64">
<dim>3</dim>
</port>
</output>
</layer>
<layer id="137" name="Constant_10812" type="Const" version="opset1">
<data element_type="i64" shape="" offset="135272" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="138" name="Constant_10813" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="139" name="Gather_10814" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I64">
<dim>3</dim>
</port>
<port id="1" precision="I64" />
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="I64" names="77" />
</output>
</layer>
<layer id="140" name="__module.model/aten::add/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="I64" />
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="85,86,89,kv_length" />
</output>
</layer>
<layer id="141" name="__module.model/aten::arange/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135292" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="142" name="__module.model/aten::arange/Range" type="Range" version="opset4">
<data output_type="i64" />
<input>
<port id="0" precision="I32" />
<port id="1" precision="I64" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I64" names="97,98,kv_indices">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="143" name="65" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" names="65" />
</output>
</layer>
<layer id="144" name="__module.model/aten::unsqueeze/Unsqueeze" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="107">
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="145" name="__module.model/aten::unsqueeze/Unsqueeze_1" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="108">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="146" name="60" type="Const" version="opset1">
<data element_type="i64" shape="" offset="135312" size="8" />
<output>
<port id="0" precision="I64" names="60" />
</output>
</layer>
<layer id="147" name="__module.model/aten::unsqueeze/Unsqueeze_2" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="109,110,kv_idx">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="148" name="__module.model/aten::arange/Constant_2" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="149" name="Constant_10833" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="150" name="Reshape_10834" type="Reshape" version="opset1">
<data special_zero="false" />
<input>
<port id="0" precision="I64" />
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="151" name="__module.model/aten::size/Constant_3" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="152" name="__module.model/aten::size/Gather_3" type="Squeeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>1</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="I64" names="92" />
</output>
</layer>
<layer id="153" name="__module.model/aten::arange/Constant_4" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135292" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="154" name="__module.model/aten::arange/Range_2" type="Range" version="opset4">
<data output_type="i64" />
<input>
<port id="0" precision="I32" />
<port id="1" precision="I64" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I64" names="95">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="155" name="__module.model/aten::arange/Constant_3" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135292" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="156" name="__module.model/aten::arange/Range_1" type="Range" version="opset4">
<data output_type="i64" />
<input>
<port id="0" precision="I64" />
<port id="1" precision="I64" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I64" names="91,cache_position">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="157" name="__module.model/aten::select/Gather" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="I64" names="93,q_offset" />
</output>
</layer>
<layer id="158" name="__module.model/aten::add/Add_3" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="96,q_indices">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="159" name="__module.model/aten::unsqueeze/Unsqueeze_3" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="103">
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="160" name="__module.model/aten::unsqueeze/Unsqueeze_4" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="104,105">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="161" name="56" type="Const" version="opset1">
<data element_type="i64" shape="" offset="135320" size="8" />
<output>
<port id="0" precision="I64" names="56" />
</output>
</layer>
<layer id="162" name="__module.model/aten::unsqueeze/Unsqueeze_5" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="106,q_idx">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="163" name="__module.model/aten::le/LessEqual" type="LessEqual" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL" names="113,114">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="164" name="__module.model/aten::__and__/BitwiseAnd" type="BitwiseAnd" version="opset13">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="BOOL" />
<port id="1" precision="BOOL">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL" names="115,result">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="165" name="__module.model/aten::to/Convert_2" type="Convert" version="opset1">
<data destination_type="boolean" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="BOOL" names="76,attention_mask.3">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="166" name="__module.model/aten::index/ShapeOf_1" type="ShapeOf" version="opset3">
<data output_type="i32" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>2</dim>
</port>
</output>
</layer>
<layer id="167" name="Constant_1394" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="168" name="__module.model/aten::index/ReduceProd" type="ReduceProd" version="opset1">
<data keep_dims="true" />
<input>
<port id="0" precision="I32">
<dim>2</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="I32">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="169" name="Constant_1396" type="Const" version="opset1">
<data element_type="i32" shape="1" offset="135328" size="4" />
<output>
<port id="0" precision="I32">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="170" name="__module.model/aten::index/Concat" type="Concat" version="opset1">
<data axis="0" />
<input>
<port id="0" precision="I32">
<dim>1</dim>
</port>
<port id="1" precision="I32">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>2</dim>
</port>
</output>
</layer>
<layer id="171" name="__module.model/aten::index/Reshape" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I32">
<dim>2</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL">
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="172" name="__module.model/aten::index/Convert" type="Convert" version="opset1">
<data destination_type="i32" />
<input>
<port id="0" precision="I64">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="173" name="__module.model/aten::arange/Constant_5" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="174" name="Constant_10816" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="175" name="Constant_10817" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="176" name="Gather_10818" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="I64">
<dim>3</dim>
</port>
<port id="1" precision="I64" />
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="I64" names="88" />
</output>
</layer>
<layer id="177" name="__module.model/aten::arange/Constant_6" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135292" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="178" name="__module.model/aten::arange/Range_3" type="Range" version="opset4">
<data output_type="i64" />
<input>
<port id="0" precision="I32" />
<port id="1" precision="I64" />
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="I64" names="94,99,batch_indices">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="179" name="__module.model/aten::unsqueeze/Unsqueeze_6" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="100">
<dim>-1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="180" name="__module.model/aten::unsqueeze/Unsqueeze_7" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="101">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="181" name="__module.model/aten::unsqueeze/Unsqueeze_8" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
<port id="1" precision="I64" />
</input>
<output>
<port id="2" precision="I64" names="102,batch_idx">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="182" name="__module.model/aten::index/Convert_3" type="Convert" version="opset1">
<data destination_type="i32" />
<input>
<port id="0" precision="I64">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="183" name="__module.model/aten::index/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="184" name="__module.model/aten::index/Split" type="Split" version="opset1">
<data num_splits="2" />
<input>
<port id="0" precision="I32">
<dim>2</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="I32">
<dim>1</dim>
</port>
<port id="3" precision="I32">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="185" name="__module.model/aten::index/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
<port id="1" precision="I32">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="186" name="__module.model/aten::index/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="I32">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="187" name="__module.model/aten::index/Gather" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
<port id="2" precision="I32" />
</input>
<output>
<port id="3" precision="BOOL">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="188" name="Constant_5033" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="131104" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="189" name="__module.model/aten::index/Reshape_8" type="Reshape" version="opset1">
<data special_zero="false" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL">
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="190" name="__module.model/aten::index/ShapeOf_5" type="ShapeOf" version="opset3">
<data output_type="i32" />
<input>
<port id="0" precision="I32">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="191" name="__module.model/aten::index/Reshape_12" type="Reshape" version="opset1">
<data special_zero="false" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL" names="117,118">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="192" name="__module.model/aten::__and__/BitwiseAnd_1" type="BitwiseAnd" version="opset13">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="BOOL">
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="BOOL">
<dim>-1</dim>
<dim>1</dim>
<dim>1</dim>
<dim>-1</dim>
</port>
</input>
<output>
<port id="2" precision="BOOL" names="119,attention_mask.5">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="193" name="Constant_10835" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="194" name="Reshape_10836" type="Reshape" version="opset1">
<data special_zero="false" />
<input>
<port id="0" precision="I64" />
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="195" name="Constant_8931" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="196" name="Constant_175" type="Const" version="opset1">
<data element_type="i32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="197" name="Unsqueeze_179" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="I64" />
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="198" name="__module.model/prim::ListConstruct/SequenceMark" 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>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="4" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="199" name="__module.model/aten::expand/Broadcast" type="Broadcast" version="opset3">
<data mode="bidirectional" />
<input>
<port id="0" precision="BOOL">
<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="BOOL" names="121,mask">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="200" name="__module.model/aten::to/Convert_4" type="Const" version="opset1">
<data element_type="f32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="FP32" names="122" />
</output>
</layer>
<layer id="201" name="__module.model/aten::to/Convert_5" type="Const" version="opset1">
<data element_type="f32" shape="" offset="143564" size="4" />
<output>
<port id="0" precision="FP32" names="124" />
</output>
</layer>
<layer id="202" name="__module.model/aten::where/Select" type="Select" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="BOOL">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="1" precision="FP32" />
<port id="2" precision="FP32" />
</input>
<output>
<port id="3" precision="FP32" names="126,attention_mask_1">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
</output>
</layer>
<layer id="203" name="__module.model.layers.0.self_attn/aten::scaled_dot_product_attention/ConvertLike" type="Const" version="opset1">
<data element_type="f32" shape="" offset="143568" size="4" />
<output>
<port id="0" precision="FP32" />
</output>
</layer>
<layer id="204" name="__module.model.layers.0.self_attn/aten::scaled_dot_product_attention/ScaledDotProductAttention" type="ScaledDotProductAttention" version="opset13">
<data causal="false" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="4" precision="FP32" />
</input>
<output>
<port id="5" precision="FP32" names="223,attn_output.1">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="205" name="__module.model.layers.0.self_attn/aten::transpose/Constant_3" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="206" name="__module.model.layers.0.self_attn/aten::transpose/Transpose_3" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="224">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="207" name="Constant_10868" type="Const" version="opset1">
<data element_type="i64" shape="3" offset="143576" size="24" />
<output>
<port id="0" precision="I64">
<dim>3</dim>
</port>
</output>
</layer>
<layer id="208" name="__module.model.layers.0.self_attn/aten::reshape/Reshape" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>3</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="227">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="209" name="self.model.layers.0.self_attn.o_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="143600" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.self_attn.o_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="210" name="__module.model.layers.0.self_attn.o_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="230,hidden_states_attention.1">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="211" name="__module.model.layers.0/aten::add/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="246">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="212" name="self.model.layers.0.mlp.gate_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="37, 32" offset="147696" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.mlp.gate_proj.weight">
<dim>37</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="213" name="__module.model.layers.0.mlp.gate_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>37</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="239,input.3">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="214" name="__module.model.layers.0.mlp.act_fn/aten::gelu/Gelu" type="Gelu" version="opset7">
<data approximation_mode="ERF" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="240">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="215" name="self.model.layers.0.mlp.up_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="37, 32" offset="152432" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.mlp.up_proj.weight">
<dim>37</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="216" name="__module.model.layers.0.mlp.up_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>37</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="242">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="217" name="__module.model.layers.0.mlp/aten::mul/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="243,input.5">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="218" name="self.model.layers.0.mlp.down_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 37" offset="157168" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.0.mlp.down_proj.weight">
<dim>32</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="219" name="__module.model.layers.0.mlp.down_proj/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>37</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="245,hidden_states_mlp.1">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="220" name="__module.model.layers.0/aten::add/Add_1" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="247,256,hidden_states.13,hidden_states.15">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="221" name="Constant_780" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="131104" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="222" name="__module.model.layers.1.input_layernorm/aten::mul/Multiply" type="MVN" version="opset6">
<data eps="9.9999997473787516e-06" normalize_variance="true" eps_mode="INSIDE_SQRT" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="266,268,269,hidden_states.17,hidden_states.19,hidden_states.21">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="223" name="self.model.layers.1.self_attn.v_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="161904" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.self_attn.v_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="224" name="__module.model.layers.1.self_attn.v_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="288">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="225" name="Constant_10869" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="226" name="__module.model.layers.1.self_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>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="290,value_states.5">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="227" name="__module.model.layers.1.self_attn/aten::transpose/Constant" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="228" name="__module.model.layers.1.self_attn/aten::transpose/Transpose" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="293,value_states">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="229" name="__module.model.layers.1.self_attn/aten::cat/Concat" type="Concat" version="opset1">
<data axis="-2" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="hidden_states.25,present.1.value">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="230" name="Constant_12308" type="Const" version="opset1">
<data element_type="f32" shape="" offset="0" size="4" />
<output>
<port id="0" precision="FP32" />
</output>
</layer>
<layer id="231" name="Constant_12301" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="16" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="232" name="Constant_12303" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="233" name="Constant_12305" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="24" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="234" name="Concat_12306" 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>
<port id="3" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="4" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="235" name="Broadcast_12309" type="Broadcast" version="opset3">
<data mode="numpy" />
<input>
<port id="0" precision="FP32" />
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="236" name="ReadValue_12087" type="ReadValue" version="opset6">
<data variable_id="past_key_values.1.keypresent.1.key" variable_type="f32" variable_shape="?,4,?,8" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>0</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="past_key_values.1.key">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="237" name="Constant_11911" type="Const" version="opset1">
<data element_type="i64" shape="" offset="8" size="8" />
<output>
<port id="0" precision="I64" />
</output>
</layer>
<layer id="238" name="Gather_11912" type="Gather" version="opset8">
<data batch_dims="0" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>-1</dim>
</port>
<port id="2" precision="I64" />
</input>
<output>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="239" name="self.model.layers.1.self_attn.k_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="166000" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.self_attn.k_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="240" name="__module.model.layers.1.self_attn.k_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="284">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="241" name="Constant_10870" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="242" name="__module.model.layers.1.self_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>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="286,key_states.5">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="243" name="__module.model.layers.1.self_attn/aten::transpose/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="244" name="__module.model.layers.1.self_attn/aten::to/Convert" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="295,k">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="245" name="__module.model.layers.1.self_attn/aten::mul/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="307">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="246" name="Constant_1066" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="247" name="Constant_1068" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="248" name="Constant_1070" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="249" name="Constant_1064" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="250" name="__module.model.layers.1.self_attn/aten::slice/Slice" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="309,x2">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="251" name="__module.model.layers.1.self_attn/aten::neg/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="252" name="__module.model.layers.1.self_attn/aten::neg/ConvertLike" type="Convert" version="opset1">
<data destination_type="f32" />
<input>
<port id="0" precision="I32" />
</input>
<output>
<port id="1" precision="FP32" />
</output>
</layer>
<layer id="253" name="__module.model.layers.1.self_attn/aten::neg/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="FP32" />
</input>
<output>
<port id="2" precision="FP32" names="310">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="254" name="__module.model.layers.1.self_attn/aten::stack/Constant" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="255" name="__module.model.layers.1.self_attn/aten::stack/Unsqueeze" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="256" name="Constant_1056" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="257" name="Constant_1058" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="258" name="Constant_1060" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="259" name="Constant_1054" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="260" name="__module.model.layers.1.self_attn/aten::slice/Slice_1" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="308,x1">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="261" name="__module.model.layers.1.self_attn/aten::stack/Unsqueeze_1" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="262" name="__module.model.layers.1.self_attn/aten::stack/Concat" type="Concat" version="opset1">
<data axis="-1" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="312">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
</output>
</layer>
<layer id="263" name="Constant_10871" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135336" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="264" name="__module.model.layers.1.self_attn/aten::flatten/Reshape" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="313">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="265" name="__module.model.layers.1.self_attn/aten::mul/Multiply_1" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="314">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="266" name="__module.model.layers.1.self_attn/aten::add/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="315,317,k_embed,key_states">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="267" name="__module.model.layers.1.self_attn/aten::cat/Concat_1" type="Concat" version="opset1">
<data axis="-2" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="hidden_states.23">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="268" name="self.model.layers.1.self_attn.q_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="170096" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.self_attn.q_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="269" name="__module.model.layers.1.self_attn.q_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="280">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="270" name="Constant_10872" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135208" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="271" name="__module.model.layers.1.self_attn/aten::view/Reshape_2" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="282,query_states">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="272" name="__module.model.layers.1.self_attn/aten::transpose/Constant_2" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="273" name="__module.model.layers.1.self_attn/aten::to/Convert_2" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="294,q">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="274" name="__module.model.layers.1.self_attn/aten::mul/Multiply_2" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="298">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="275" name="Constant_1012" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135272" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="276" name="Constant_1014" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="277" name="Constant_1016" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="278" name="Constant_1010" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="279" name="__module.model.layers.1.self_attn/aten::slice/Slice_2" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="300,x2.5">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="280" name="__module.model.layers.1.self_attn/aten::neg/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="281" name="__module.model.layers.1.self_attn/aten::neg/ConvertLike_1" type="Convert" version="opset1">
<data destination_type="f32" />
<input>
<port id="0" precision="I32" />
</input>
<output>
<port id="1" precision="FP32" />
</output>
</layer>
<layer id="282" name="__module.model.layers.1.self_attn/aten::neg/Multiply_1" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="FP32" />
</input>
<output>
<port id="2" precision="FP32" names="301">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="283" name="__module.model.layers.1.self_attn/aten::stack/Constant_1" type="Const" version="opset1">
<data element_type="i32" shape="" offset="135328" size="4" />
<output>
<port id="0" precision="I32" />
</output>
</layer>
<layer id="284" name="__module.model.layers.1.self_attn/aten::stack/Unsqueeze_2" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="285" name="Constant_1002" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="8" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="286" name="Constant_1004" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135304" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="287" name="Constant_1006" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135312" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="288" name="Constant_1000" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="135320" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="289" name="__module.model.layers.1.self_attn/aten::slice/Slice_3" type="Slice" version="opset8">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
<port id="2" precision="I64">
<dim>1</dim>
</port>
<port id="3" precision="I64">
<dim>1</dim>
</port>
<port id="4" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="5" precision="FP32" names="299,x1.5">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
</output>
</layer>
<layer id="290" name="__module.model.layers.1.self_attn/aten::stack/Unsqueeze_3" type="Unsqueeze" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
</port>
<port id="1" precision="I32" />
</input>
<output>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="291" name="__module.model.layers.1.self_attn/aten::stack/Concat_1" type="Concat" version="opset1">
<data axis="-1" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="303">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
</output>
</layer>
<layer id="292" name="Constant_10873" type="Const" version="opset1">
<data element_type="i64" shape="4" offset="135336" size="32" />
<output>
<port id="0" precision="I64">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="293" name="__module.model.layers.1.self_attn/aten::flatten/Reshape_1" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>2</dim>
</port>
<port id="1" precision="I64">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="304">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="294" name="__module.model.layers.1.self_attn/aten::mul/Multiply_3" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="305">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="295" name="__module.model.layers.1.self_attn/aten::add/Add_1" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="306,316,q_embed,query">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="296" name="__module.model.layers.1.self_attn/aten::scaled_dot_product_attention/ScaledDotProductAttention" type="ScaledDotProductAttention" version="opset13">
<data causal="false" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="2" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="3" precision="FP32">
<dim>-1</dim>
<dim>1</dim>
<dim>-1</dim>
<dim>-1</dim>
</port>
<port id="4" precision="FP32" />
</input>
<output>
<port id="5" precision="FP32" names="322,attn_output.5">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="297" name="__module.model.layers.1.self_attn/aten::transpose/Constant_3" type="Const" version="opset1">
<data element_type="i32" shape="4" offset="135240" size="16" />
<output>
<port id="0" precision="I32">
<dim>4</dim>
</port>
</output>
</layer>
<layer id="298" name="__module.model.layers.1.self_attn/aten::transpose/Transpose_3" type="Transpose" version="opset1">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
<port id="1" precision="I32">
<dim>4</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="323">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="299" name="Constant_10874" type="Const" version="opset1">
<data element_type="i64" shape="3" offset="143576" size="24" />
<output>
<port id="0" precision="I64">
<dim>3</dim>
</port>
</output>
</layer>
<layer id="300" name="__module.model.layers.1.self_attn/aten::reshape/Reshape" type="Reshape" version="opset1">
<data special_zero="true" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>4</dim>
<dim>8</dim>
</port>
<port id="1" precision="I64">
<dim>3</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="326">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="301" name="self.model.layers.1.self_attn.o_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 32" offset="174192" size="4096" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.self_attn.o_proj.weight">
<dim>32</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="302" name="__module.model.layers.1.self_attn.o_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="329,hidden_states_attention">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="303" name="__module.model.layers.1/aten::add/Add" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="345">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="304" name="self.model.layers.1.mlp.gate_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="37, 32" offset="178288" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.mlp.gate_proj.weight">
<dim>37</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="305" name="__module.model.layers.1.mlp.gate_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>37</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="338,input.9">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="306" name="__module.model.layers.1.mlp.act_fn/aten::gelu/Gelu" type="Gelu" version="opset7">
<data approximation_mode="ERF" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="1" precision="FP32" names="339">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="307" name="self.model.layers.1.mlp.up_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="37, 32" offset="183024" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.mlp.up_proj.weight">
<dim>37</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="308" name="__module.model.layers.1.mlp.up_proj/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>37</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="341">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="309" name="__module.model.layers.1.mlp/aten::mul/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="342,input.11">
<dim>-1</dim>
<dim>-1</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="310" name="self.model.layers.1.mlp.down_proj.weight" type="Const" version="opset1">
<data element_type="f32" shape="32, 37" offset="187760" size="4736" />
<output>
<port id="0" precision="FP32" names="self.model.layers.1.mlp.down_proj.weight">
<dim>32</dim>
<dim>37</dim>
</port>
</output>
</layer>
<layer id="311" name="__module.model.layers.1.mlp.down_proj/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>37</dim>
</port>
<port id="1" precision="FP32">
<dim>32</dim>
<dim>37</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="344,hidden_states_mlp">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="312" name="__module.model.layers.1/aten::add/Add_1" type="Add" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="346,352,hidden_states.27,hidden_states.29">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="313" name="Constant_1204" type="Const" version="opset1">
<data element_type="i64" shape="1" offset="131104" size="8" />
<output>
<port id="0" precision="I64">
<dim>1</dim>
</port>
</output>
</layer>
<layer id="314" name="__module.model.norm/aten::mul/Multiply" type="MVN" version="opset6">
<data eps="9.9999997473787516e-06" normalize_variance="true" eps_mode="INSIDE_SQRT" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
<port id="1" precision="I64">
<dim>1</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="26,31,36,362,364,365,hidden_states,hidden_states.31,hidden_states.33,input">
<dim>-1</dim>
<dim>-1</dim>
<dim>32</dim>
</port>
</output>
</layer>
<layer id="315" name="__module.lm_head/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>32</dim>
</port>
<port id="1" precision="FP32">
<dim>1024</dim>
<dim>32</dim>
</port>
</input>
<output>
<port id="2" precision="FP32" names="368,logits_1">
<dim>-1</dim>
<dim>-1</dim>
<dim>1024</dim>
</port>
</output>
</layer>
<layer id="316" name="Constant_10524" type="Const" version="opset1">
<data element_type="f32" shape="1, 1, 1" offset="192496" size="4" />
<output>
<port id="0" precision="FP32">
<dim>1</dim>
<dim>1</dim>
<dim>1</dim>
</port>
</output>
</layer>
<layer id="317" name="aten::mul/Multiply" type="Multiply" version="opset1">
<data auto_broadcast="numpy" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>1024</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="logits">
<dim>-1</dim>
<dim>-1</dim>
<dim>1024</dim>
</port>
</output>
</layer>
<layer id="318" name="Assign_12084" type="Assign" version="opset6">
<data variable_id="past_key_values.0.keypresent.0.key" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="319" name="Assign_12086" type="Assign" version="opset6">
<data variable_id="past_key_values.0.valuepresent.0.value" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="320" name="Assign_12088" type="Assign" version="opset6">
<data variable_id="past_key_values.1.keypresent.1.key" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="321" name="Assign_12090" type="Assign" version="opset6">
<data variable_id="past_key_values.1.valuepresent.1.value" />
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</input>
<output>
<port id="1" precision="FP32">
<dim>-1</dim>
<dim>4</dim>
<dim>-1</dim>
<dim>8</dim>
</port>
</output>
</layer>
<layer id="322" name="Result_1444" type="Result" version="opset1" output_names="logits">
<input>
<port id="0" precision="FP32">
<dim>-1</dim>
<dim>-1</dim>
<dim>1024</dim>
</port>
</input>
</layer>
</layers>
<edges>
<edge from-layer="0" from-port="0" to-layer="5" to-port="0" />
<edge from-layer="0" from-port="0" to-layer="18" to-port="0" />
<edge from-layer="1" from-port="0" to-layer="165" to-port="0" />
<edge from-layer="2" from-port="0" to-layer="31" to-port="0" />
<edge from-layer="3" from-port="0" to-layer="122" to-port="1" />
<edge from-layer="3" from-port="0" to-layer="238" to-port="1" />
<edge from-layer="3" from-port="0" to-layer="84" to-port="1" />
<edge from-layer="3" from-port="0" to-layer="16" to-port="1" />
<edge from-layer="4" from-port="0" to-layer="13" to-port="0" />
<edge from-layer="5" from-port="1" to-layer="8" to-port="0" />
<edge from-layer="6" from-port="0" to-layer="8" to-port="1" />
<edge from-layer="7" from-port="0" to-layer="8" to-port="2" />
<edge from-layer="8" from-port="3" to-layer="80" to-port="0" />
<edge from-layer="8" from-port="3" to-layer="234" to-port="0" />
<edge from-layer="8" from-port="3" to-layer="12" to-port="0" />
<edge from-layer="8" from-port="3" to-layer="118" to-port="0" />
<edge from-layer="9" from-port="0" to-layer="12" to-port="1" />
<edge from-layer="10" from-port="0" to-layer="12" to-port="2" />
<edge from-layer="11" from-port="0" to-layer="12" to-port="3" />
<edge from-layer="12" from-port="4" to-layer="13" to-port="1" />
<edge from-layer="13" from-port="2" to-layer="14" to-port="0" />
<edge from-layer="14" from-port="1" to-layer="16" to-port="0" />
<edge from-layer="15" from-port="0" to-layer="16" to-port="2" />
<edge from-layer="16" from-port="3" to-layer="229" to-port="0" />
<edge from-layer="17" from-port="0" to-layer="315" to-port="1" />
<edge from-layer="17" from-port="0" to-layer="20" to-port="0" />
<edge from-layer="18" from-port="1" to-layer="20" to-port="1" />
<edge from-layer="19" from-port="0" to-layer="20" to-port="2" />
<edge from-layer="20" from-port="3" to-layer="22" to-port="0" />
<edge from-layer="20" from-port="3" to-layer="211" to-port="0" />
<edge from-layer="21" from-port="0" to-layer="22" to-port="1" />
<edge from-layer="22" from-port="2" to-layer="24" to-port="0" />
<edge from-layer="22" from-port="2" to-layer="136" to-port="0" />
<edge from-layer="22" from-port="2" to-layer="124" to-port="0" />
<edge from-layer="22" from-port="2" to-layer="216" to-port="0" />
<edge from-layer="22" from-port="2" to-layer="213" to-port="0" />
<edge from-layer="22" from-port="2" to-layer="86" to-port="0" />
<edge from-layer="23" from-port="0" to-layer="24" to-port="1" />
<edge from-layer="24" from-port="2" to-layer="26" to-port="0" />
<edge from-layer="25" from-port="0" to-layer="26" to-port="1" />
<edge from-layer="26" from-port="2" to-layer="28" to-port="0" />
<edge from-layer="27" from-port="0" to-layer="28" to-port="1" />
<edge from-layer="28" from-port="2" to-layer="57" to-port="0" />
<edge from-layer="28" from-port="2" to-layer="67" to-port="0" />
<edge from-layer="28" from-port="2" to-layer="52" to-port="0" />
<edge from-layer="29" from-port="0" to-layer="33" to-port="0" />
<edge from-layer="30" from-port="0" to-layer="31" to-port="1" />
<edge from-layer="30" from-port="0" to-layer="160" to-port="1" />
<edge from-layer="30" from-port="0" to-layer="145" to-port="1" />
<edge from-layer="30" from-port="0" to-layer="179" to-port="1" />
<edge from-layer="30" from-port="0" to-layer="73" to-port="1" />
<edge from-layer="30" from-port="0" to-layer="51" to-port="1" />
<edge from-layer="31" from-port="2" to-layer="32" to-port="0" />
<edge from-layer="32" from-port="1" 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="49" to-port="0" />
<edge from-layer="35" from-port="2" to-layer="37" to-port="0" />
<edge from-layer="36" from-port="0" to-layer="42" to-port="1" />
<edge from-layer="36" from-port="0" to-layer="39" to-port="2" />
<edge from-layer="36" from-port="0" to-layer="41" to-port="0" />
<edge from-layer="37" from-port="1" to-layer="39" to-port="0" />
<edge from-layer="38" from-port="0" to-layer="39" to-port="1" />
<edge from-layer="38" from-port="0" to-layer="49" to-port="2" />
<edge from-layer="39" from-port="3" to-layer="41" to-port="1" />
<edge from-layer="40" from-port="0" to-layer="41" to-port="2" />
<edge from-layer="41" from-port="3" to-layer="42" to-port="0" />
<edge from-layer="42" from-port="2" to-layer="44" to-port="0" />
<edge from-layer="43" from-port="0" to-layer="44" to-port="1" />
<edge from-layer="44" from-port="2" to-layer="46" to-port="0" />
<edge from-layer="45" from-port="0" to-layer="46" to-port="1" />
<edge from-layer="46" from-port="2" to-layer="48" to-port="0" />
<edge from-layer="47" from-port="0" to-layer="48" to-port="1" />
<edge from-layer="48" from-port="2" to-layer="49" to-port="1" />
<edge from-layer="49" from-port="3" to-layer="72" to-port="0" />
<edge from-layer="49" from-port="3" to-layer="50" to-port="0" />
<edge from-layer="50" from-port="1" to-layer="51" to-port="0" />
<edge from-layer="51" from-port="2" to-layer="91" to-port="1" />
<edge from-layer="51" from-port="2" to-layer="274" to-port="1" />
<edge from-layer="51" from-port="2" to-layer="52" to-port="1" />
<edge from-layer="51" from-port="2" to-layer="245" to-port="1" />
<edge from-layer="52" from-port="2" to-layer="75" to-port="0" />
<edge from-layer="53" from-port="0" to-layer="57" to-port="1" />
<edge from-layer="54" from-port="0" to-layer="57" to-port="2" />
<edge from-layer="55" from-port="0" to-layer="57" to-port="3" />
<edge from-layer="56" from-port="0" to-layer="57" to-port="4" />
<edge from-layer="57" from-port="5" to-layer="60" to-port="0" />
<edge from-layer="58" from-port="0" to-layer="59" to-port="0" />
<edge from-layer="59" from-port="1" 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="68" to-port="1" />
<edge from-layer="61" from-port="0" to-layer="62" to-port="1" />
<edge from-layer="62" from-port="2" to-layer="69" to-port="0" />
<edge from-layer="63" from-port="0" to-layer="67" to-port="1" />
<edge from-layer="64" from-port="0" to-layer="67" to-port="2" />
<edge from-layer="65" from-port="0" to-layer="67" to-port="3" />
<edge from-layer="66" from-port="0" to-layer="67" to-port="4" />
<edge from-layer="67" from-port="5" to-layer="68" to-port="0" />
<edge from-layer="68" from-port="2" to-layer="69" to-port="1" />
<edge from-layer="69" from-port="2" 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="74" to-port="0" />
<edge from-layer="72" from-port="1" to-layer="73" to-port="0" />
<edge from-layer="73" from-port="2" to-layer="294" to-port="1" />
<edge from-layer="73" from-port="2" to-layer="74" to-port="1" />
<edge from-layer="73" from-port="2" to-layer="265" to-port="1" />
<edge from-layer="73" from-port="2" to-layer="111" to-port="1" />
<edge from-layer="74" from-port="2" to-layer="75" to-port="1" />
<edge from-layer="75" from-port="2" to-layer="204" to-port="0" />
<edge from-layer="76" from-port="0" to-layer="81" to-port="0" />
<edge from-layer="77" from-port="0" to-layer="80" to-port="1" />
<edge from-layer="78" from-port="0" to-layer="80" to-port="2" />
<edge from-layer="79" from-port="0" to-layer="80" to-port="3" />
<edge from-layer="80" from-port="4" to-layer="81" to-port="1" />
<edge from-layer="81" from-port="2" to-layer="82" to-port="0" />
<edge from-layer="82" from-port="1" to-layer="84" to-port="0" />
<edge from-layer="83" from-port="0" to-layer="84" to-port="2" />
<edge from-layer="84" from-port="3" to-layer="132" to-port="0" />
<edge from-layer="84" from-port="3" to-layer="113" to-port="0" />
<edge from-layer="85" from-port="0" to-layer="86" to-port="1" />
<edge from-layer="86" from-port="2" 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="90" to-port="0" />
<edge from-layer="89" from-port="0" to-layer="90" to-port="1" />
<edge from-layer="90" from-port="2" to-layer="106" to-port="0" />
<edge from-layer="90" from-port="2" to-layer="91" to-port="0" />
<edge from-layer="90" from-port="2" to-layer="96" to-port="0" />
<edge from-layer="91" from-port="2" to-layer="112" to-port="0" />
<edge from-layer="92" from-port="0" to-layer="96" to-port="1" />
<edge from-layer="93" from-port="0" to-layer="96" to-port="2" />
<edge from-layer="94" from-port="0" to-layer="96" to-port="3" />
<edge from-layer="95" from-port="0" to-layer="96" to-port="4" />
<edge from-layer="96" from-port="5" to-layer="99" to-port="0" />
<edge from-layer="97" from-port="0" to-layer="98" to-port="0" />
<edge from-layer="98" from-port="1" to-layer="99" to-port="1" />
<edge from-layer="99" from-port="2" to-layer="101" to-port="0" />
<edge from-layer="100" from-port="0" to-layer="101" to-port="1" />
<edge from-layer="100" from-port="0" to-layer="107" to-port="1" />
<edge from-layer="101" from-port="2" to-layer="108" to-port="0" />
<edge from-layer="102" from-port="0" to-layer="106" to-port="1" />
<edge from-layer="103" from-port="0" to-layer="106" to-port="2" />
<edge from-layer="104" from-port="0" to-layer="106" to-port="3" />
<edge from-layer="105" from-port="0" to-layer="106" to-port="4" />
<edge from-layer="106" from-port="5" to-layer="107" to-port="0" />
<edge from-layer="107" from-port="2" to-layer="108" to-port="1" />
<edge from-layer="108" from-port="2" to-layer="110" to-port="0" />
<edge from-layer="109" from-port="0" to-layer="110" to-port="1" />
<edge from-layer="110" from-port="2" to-layer="111" to-port="0" />
<edge from-layer="111" from-port="2" to-layer="112" to-port="1" />
<edge from-layer="112" from-port="2" to-layer="113" to-port="1" />
<edge from-layer="113" from-port="2" to-layer="204" to-port="1" />
<edge from-layer="113" from-port="2" to-layer="318" to-port="0" />
<edge from-layer="114" from-port="0" to-layer="119" to-port="0" />
<edge from-layer="115" from-port="0" to-layer="118" to-port="1" />
<edge from-layer="116" from-port="0" to-layer="118" to-port="2" />
<edge from-layer="117" from-port="0" to-layer="118" to-port="3" />
<edge from-layer="118" from-port="4" to-layer="119" to-port="1" />
<edge from-layer="119" from-port="2" to-layer="120" to-port="0" />
<edge from-layer="120" from-port="1" to-layer="122" to-port="0" />
<edge from-layer="121" from-port="0" to-layer="122" to-port="2" />
<edge from-layer="122" from-port="3" to-layer="129" to-port="0" />
<edge from-layer="123" from-port="0" to-layer="124" to-port="1" />
<edge from-layer="124" from-port="2" to-layer="126" to-port="0" />
<edge from-layer="125" from-port="0" to-layer="126" to-port="1" />
<edge from-layer="126" from-port="2" to-layer="128" to-port="0" />
<edge from-layer="127" from-port="0" to-layer="128" to-port="1" />
<edge from-layer="128" from-port="2" to-layer="129" to-port="1" />
<edge from-layer="129" from-port="2" to-layer="204" to-port="2" />
<edge from-layer="129" from-port="2" to-layer="319" to-port="0" />
<edge from-layer="130" from-port="0" to-layer="164" to-port="0" />
<edge from-layer="131" from-port="0" to-layer="142" to-port="0" />
<edge from-layer="132" from-port="1" to-layer="135" to-port="0" />
<edge from-layer="133" from-port="0" to-layer="135" to-port="1" />
<edge from-layer="134" from-port="0" to-layer="135" to-port="2" />
<edge from-layer="135" from-port="3" to-layer="156" to-port="0" />
<edge from-layer="135" from-port="3" to-layer="140" to-port="0" />
<edge from-layer="136" from-port="1" to-layer="139" to-port="0" />
<edge from-layer="136" from-port="1" to-layer="176" to-port="0" />
<edge from-layer="137" from-port="0" to-layer="139" to-port="1" />
<edge from-layer="138" from-port="0" to-layer="139" to-port="2" />
<edge from-layer="139" from-port="3" to-layer="150" to-port="0" />
<edge from-layer="139" from-port="3" to-layer="140" to-port="1" />
<edge from-layer="140" from-port="2" to-layer="156" to-port="1" />
<edge from-layer="140" from-port="2" to-layer="142" to-port="1" />
<edge from-layer="140" from-port="2" to-layer="197" to-port="0" />
<edge from-layer="141" from-port="0" to-layer="142" to-port="2" />
<edge from-layer="142" from-port="3" to-layer="144" to-port="0" />
<edge from-layer="143" from-port="0" to-layer="157" to-port="2" />
<edge from-layer="143" from-port="0" to-layer="144" to-port="1" />
<edge from-layer="143" from-port="0" to-layer="159" to-port="1" />
<edge from-layer="143" from-port="0" to-layer="157" to-port="1" />
<edge from-layer="144" from-port="2" to-layer="145" to-port="0" />
<edge from-layer="145" from-port="2" to-layer="147" to-port="0" />
<edge from-layer="146" from-port="0" to-layer="147" to-port="1" />
<edge from-layer="146" from-port="0" to-layer="180" to-port="1" />
<edge from-layer="147" from-port="2" to-layer="163" to-port="0" />
<edge from-layer="147" from-port="2" to-layer="172" to-port="0" />
<edge from-layer="148" from-port="0" to-layer="154" to-port="0" />
<edge from-layer="149" from-port="0" to-layer="150" to-port="1" />
<edge from-layer="150" from-port="2" to-layer="152" to-port="0" />
<edge from-layer="150" from-port="2" to-layer="198" to-port="2" />
<edge from-layer="151" from-port="0" to-layer="152" to-port="1" />
<edge from-layer="152" from-port="2" to-layer="154" to-port="1" />
<edge from-layer="153" from-port="0" to-layer="154" to-port="2" />
<edge from-layer="154" from-port="3" to-layer="158" to-port="0" />
<edge from-layer="155" from-port="0" to-layer="156" to-port="2" />
<edge from-layer="156" from-port="3" to-layer="157" to-port="0" />
<edge from-layer="157" from-port="3" to-layer="158" to-port="1" />
<edge from-layer="158" from-port="2" to-layer="159" to-port="0" />
<edge from-layer="159" from-port="2" to-layer="160" to-port="0" />
<edge from-layer="160" from-port="2" to-layer="162" to-port="0" />
<edge from-layer="161" from-port="0" to-layer="181" to-port="1" />
<edge from-layer="161" from-port="0" to-layer="162" to-port="1" />
<edge from-layer="162" from-port="2" to-layer="163" to-port="1" />
<edge from-layer="163" from-port="2" to-layer="164" to-port="1" />
<edge from-layer="164" from-port="2" to-layer="192" to-port="0" />
<edge from-layer="165" from-port="1" to-layer="171" to-port="0" />
<edge from-layer="165" from-port="1" to-layer="166" to-port="0" />
<edge from-layer="166" from-port="1" to-layer="184" to-port="0" />
<edge from-layer="166" from-port="1" to-layer="168" to-port="0" />
<edge from-layer="167" from-port="0" to-layer="168" to-port="1" />
<edge from-layer="168" from-port="2" to-layer="170" to-port="0" />
<edge from-layer="169" from-port="0" to-layer="170" to-port="1" />
<edge from-layer="170" from-port="2" to-layer="171" to-port="1" />
<edge from-layer="171" from-port="2" to-layer="187" to-port="0" />
<edge from-layer="172" from-port="1" to-layer="186" to-port="0" />
<edge from-layer="173" from-port="0" to-layer="178" to-port="0" />
<edge from-layer="174" from-port="0" to-layer="176" to-port="1" />
<edge from-layer="175" from-port="0" to-layer="176" to-port="2" />
<edge from-layer="176" from-port="3" to-layer="194" to-port="0" />
<edge from-layer="176" from-port="3" to-layer="178" to-port="1" />
<edge from-layer="177" from-port="0" to-layer="178" to-port="2" />
<edge from-layer="178" from-port="3" to-layer="179" to-port="0" />
<edge from-layer="179" from-port="2" to-layer="180" to-port="0" />
<edge from-layer="180" from-port="2" to-layer="181" to-port="0" />
<edge from-layer="181" from-port="2" to-layer="182" to-port="0" />
<edge from-layer="182" from-port="1" to-layer="185" to-port="0" />
<edge from-layer="183" from-port="0" to-layer="187" to-port="2" />
<edge from-layer="183" from-port="0" to-layer="184" to-port="1" />
<edge from-layer="184" from-port="3" to-layer="185" to-port="1" />
<edge from-layer="185" from-port="2" to-layer="186" to-port="1" />
<edge from-layer="186" from-port="2" to-layer="187" to-port="1" />
<edge from-layer="186" from-port="2" to-layer="190" to-port="0" />
<edge from-layer="187" from-port="3" to-layer="189" to-port="0" />
<edge from-layer="188" from-port="0" to-layer="189" to-port="1" />
<edge from-layer="189" from-port="2" to-layer="191" to-port="0" />
<edge from-layer="190" from-port="1" to-layer="191" to-port="1" />
<edge from-layer="191" from-port="2" to-layer="192" to-port="1" />
<edge from-layer="192" from-port="2" to-layer="199" to-port="0" />
<edge from-layer="193" from-port="0" to-layer="194" to-port="1" />
<edge from-layer="194" from-port="2" to-layer="198" to-port="0" />
<edge from-layer="195" from-port="0" to-layer="198" to-port="1" />
<edge from-layer="196" from-port="0" to-layer="197" to-port="1" />
<edge from-layer="197" from-port="2" to-layer="198" to-port="3" />
<edge from-layer="198" from-port="4" to-layer="199" to-port="1" />
<edge from-layer="199" from-port="2" to-layer="202" to-port="0" />
<edge from-layer="200" from-port="0" to-layer="202" to-port="1" />
<edge from-layer="201" from-port="0" to-layer="202" to-port="2" />
<edge from-layer="202" from-port="3" to-layer="204" to-port="3" />
<edge from-layer="202" from-port="3" to-layer="296" to-port="3" />
<edge from-layer="203" from-port="0" to-layer="296" to-port="4" />
<edge from-layer="203" from-port="0" to-layer="204" to-port="4" />
<edge from-layer="204" from-port="5" to-layer="206" to-port="0" />
<edge from-layer="205" from-port="0" to-layer="206" to-port="1" />
<edge from-layer="206" from-port="2" to-layer="208" to-port="0" />
<edge from-layer="207" from-port="0" to-layer="208" to-port="1" />
<edge from-layer="208" from-port="2" to-layer="210" to-port="0" />
<edge from-layer="209" from-port="0" to-layer="210" to-port="1" />
<edge from-layer="210" from-port="2" to-layer="211" to-port="1" />
<edge from-layer="211" from-port="2" to-layer="220" to-port="0" />
<edge from-layer="212" from-port="0" to-layer="213" to-port="1" />
<edge from-layer="213" from-port="2" to-layer="214" to-port="0" />
<edge from-layer="214" from-port="1" to-layer="217" to-port="0" />
<edge from-layer="215" from-port="0" to-layer="216" to-port="1" />
<edge from-layer="216" from-port="2" to-layer="217" to-port="1" />
<edge from-layer="217" from-port="2" to-layer="219" to-port="0" />
<edge from-layer="218" from-port="0" to-layer="219" to-port="1" />
<edge from-layer="219" from-port="2" to-layer="220" to-port="1" />
<edge from-layer="220" from-port="2" to-layer="303" to-port="0" />
<edge from-layer="220" from-port="2" to-layer="222" to-port="0" />
<edge from-layer="221" from-port="0" to-layer="222" to-port="1" />
<edge from-layer="222" from-port="2" to-layer="305" to-port="0" />
<edge from-layer="222" from-port="2" to-layer="240" to-port="0" />
<edge from-layer="222" from-port="2" to-layer="269" to-port="0" />
<edge from-layer="222" from-port="2" to-layer="224" to-port="0" />
<edge from-layer="222" from-port="2" to-layer="308" to-port="0" />
<edge from-layer="223" from-port="0" to-layer="224" to-port="1" />
<edge from-layer="224" from-port="2" to-layer="226" to-port="0" />
<edge from-layer="225" from-port="0" to-layer="226" to-port="1" />
<edge from-layer="226" from-port="2" to-layer="228" to-port="0" />
<edge from-layer="227" from-port="0" to-layer="228" to-port="1" />
<edge from-layer="228" from-port="2" to-layer="229" to-port="1" />
<edge from-layer="229" from-port="2" to-layer="321" to-port="0" />
<edge from-layer="229" from-port="2" to-layer="296" to-port="2" />
<edge from-layer="230" from-port="0" to-layer="235" to-port="0" />
<edge from-layer="231" from-port="0" to-layer="234" to-port="1" />
<edge from-layer="232" from-port="0" to-layer="234" to-port="2" />
<edge from-layer="233" from-port="0" to-layer="234" to-port="3" />
<edge from-layer="234" from-port="4" to-layer="235" to-port="1" />
<edge from-layer="235" from-port="2" to-layer="236" to-port="0" />
<edge from-layer="236" from-port="1" to-layer="238" to-port="0" />
<edge from-layer="237" from-port="0" to-layer="238" to-port="2" />
<edge from-layer="238" from-port="3" to-layer="267" to-port="0" />
<edge from-layer="239" from-port="0" to-layer="240" to-port="1" />
<edge from-layer="240" from-port="2" to-layer="242" to-port="0" />
<edge from-layer="241" from-port="0" to-layer="242" to-port="1" />
<edge from-layer="242" from-port="2" to-layer="244" to-port="0" />
<edge from-layer="243" from-port="0" to-layer="244" to-port="1" />
<edge from-layer="244" from-port="2" to-layer="245" to-port="0" />
<edge from-layer="244" from-port="2" to-layer="260" to-port="0" />
<edge from-layer="244" from-port="2" to-layer="250" to-port="0" />
<edge from-layer="245" from-port="2" to-layer="266" to-port="0" />
<edge from-layer="246" from-port="0" to-layer="250" to-port="1" />
<edge from-layer="247" from-port="0" to-layer="250" to-port="2" />
<edge from-layer="248" from-port="0" to-layer="250" to-port="3" />
<edge from-layer="249" from-port="0" to-layer="250" to-port="4" />
<edge from-layer="250" from-port="5" to-layer="253" to-port="0" />
<edge from-layer="251" from-port="0" to-layer="252" to-port="0" />
<edge from-layer="252" from-port="1" to-layer="253" to-port="1" />
<edge from-layer="253" from-port="2" to-layer="255" to-port="0" />
<edge from-layer="254" from-port="0" to-layer="261" to-port="1" />
<edge from-layer="254" from-port="0" to-layer="255" to-port="1" />
<edge from-layer="255" from-port="2" to-layer="262" to-port="0" />
<edge from-layer="256" from-port="0" to-layer="260" to-port="1" />
<edge from-layer="257" from-port="0" to-layer="260" to-port="2" />
<edge from-layer="258" from-port="0" to-layer="260" to-port="3" />
<edge from-layer="259" from-port="0" to-layer="260" to-port="4" />
<edge from-layer="260" from-port="5" to-layer="261" to-port="0" />
<edge from-layer="261" from-port="2" to-layer="262" to-port="1" />
<edge from-layer="262" from-port="2" to-layer="264" to-port="0" />
<edge from-layer="263" from-port="0" to-layer="264" to-port="1" />
<edge from-layer="264" from-port="2" to-layer="265" to-port="0" />
<edge from-layer="265" from-port="2" to-layer="266" to-port="1" />
<edge from-layer="266" from-port="2" to-layer="267" to-port="1" />
<edge from-layer="267" from-port="2" to-layer="320" to-port="0" />
<edge from-layer="267" from-port="2" to-layer="296" to-port="1" />
<edge from-layer="268" from-port="0" to-layer="269" to-port="1" />
<edge from-layer="269" from-port="2" to-layer="271" to-port="0" />
<edge from-layer="270" from-port="0" to-layer="271" to-port="1" />
<edge from-layer="271" from-port="2" to-layer="273" to-port="0" />
<edge from-layer="272" from-port="0" to-layer="273" to-port="1" />
<edge from-layer="273" from-port="2" to-layer="279" to-port="0" />
<edge from-layer="273" from-port="2" to-layer="274" to-port="0" />
<edge from-layer="273" from-port="2" to-layer="289" to-port="0" />
<edge from-layer="274" from-port="2" to-layer="295" to-port="0" />
<edge from-layer="275" from-port="0" to-layer="279" to-port="1" />
<edge from-layer="276" from-port="0" to-layer="279" to-port="2" />
<edge from-layer="277" from-port="0" to-layer="279" to-port="3" />
<edge from-layer="278" from-port="0" to-layer="279" to-port="4" />
<edge from-layer="279" from-port="5" to-layer="282" to-port="0" />
<edge from-layer="280" from-port="0" to-layer="281" to-port="0" />
<edge from-layer="281" from-port="1" to-layer="282" to-port="1" />
<edge from-layer="282" from-port="2" to-layer="284" to-port="0" />
<edge from-layer="283" from-port="0" to-layer="284" to-port="1" />
<edge from-layer="283" from-port="0" to-layer="290" to-port="1" />
<edge from-layer="284" from-port="2" to-layer="291" to-port="0" />
<edge from-layer="285" from-port="0" to-layer="289" to-port="1" />
<edge from-layer="286" from-port="0" to-layer="289" to-port="2" />
<edge from-layer="287" from-port="0" to-layer="289" to-port="3" />
<edge from-layer="288" from-port="0" to-layer="289" to-port="4" />
<edge from-layer="289" from-port="5" to-layer="290" to-port="0" />
<edge from-layer="290" from-port="2" to-layer="291" to-port="1" />
<edge from-layer="291" from-port="2" to-layer="293" to-port="0" />
<edge from-layer="292" from-port="0" to-layer="293" to-port="1" />
<edge from-layer="293" from-port="2" to-layer="294" to-port="0" />
<edge from-layer="294" from-port="2" to-layer="295" to-port="1" />
<edge from-layer="295" from-port="2" to-layer="296" to-port="0" />
<edge from-layer="296" from-port="5" to-layer="298" to-port="0" />
<edge from-layer="297" from-port="0" to-layer="298" to-port="1" />
<edge from-layer="298" from-port="2" to-layer="300" to-port="0" />
<edge from-layer="299" from-port="0" to-layer="300" to-port="1" />
<edge from-layer="300" from-port="2" to-layer="302" to-port="0" />
<edge from-layer="301" from-port="0" to-layer="302" to-port="1" />
<edge from-layer="302" from-port="2" to-layer="303" to-port="1" />
<edge from-layer="303" from-port="2" to-layer="312" to-port="0" />
<edge from-layer="304" from-port="0" to-layer="305" to-port="1" />
<edge from-layer="305" from-port="2" to-layer="306" to-port="0" />
<edge from-layer="306" from-port="1" to-layer="309" to-port="0" />
<edge from-layer="307" from-port="0" to-layer="308" to-port="1" />
<edge from-layer="308" from-port="2" to-layer="309" to-port="1" />
<edge from-layer="309" from-port="2" to-layer="311" to-port="0" />
<edge from-layer="310" from-port="0" to-layer="311" to-port="1" />
<edge from-layer="311" from-port="2" to-layer="312" to-port="1" />
<edge from-layer="312" from-port="2" to-layer="314" to-port="0" />
<edge from-layer="313" from-port="0" to-layer="314" to-port="1" />
<edge from-layer="314" from-port="2" to-layer="315" to-port="0" />
<edge from-layer="315" from-port="2" to-layer="317" to-port="0" />
<edge from-layer="316" from-port="0" to-layer="317" to-port="1" />
<edge from-layer="317" from-port="2" to-layer="322" to-port="0" />
</edges>
<rt_info>
<info name="OpenVINO Runtime" value="2026.5.0-23342-dc4898fd814" />
<Runtime_version value="2026.5.0-23342-dc4898fd814" />
<conversion_parameters>
<framework value="pytorch" />
<is_python_object value="True" />
</conversion_parameters>
<optimum>
<optimum_intel_version value="2.3.0.dev0+9300d0ff" />
<optimum_version value="2.3.0" />
<pytorch_version value="2.14.1+cpu" />
<transformers_version value="5.10.4" />
</optimum>
<runtime_options>
<ACTIVATIONS_SCALE_FACTOR value="8.0" />
<KV_CACHE_PRECISION value="f16" />
</runtime_options>
</rt_info>
</net>