strict digraph { "0 /nncf_model_input_0"; "1 /nncf_model_input_1"; "2 GPT2LMHeadModel/GPT2Model[transformer]/view_0"; "3 GPT2LMHeadModel/GPT2Model[transformer]/view_1"; "4 GPT2LMHeadModel/GPT2Model[transformer]/__getitem___0"; "5 GPT2LMHeadModel/GPT2Model[transformer]/NNCFEmbedding[wte]/embedding_0"; "6 GPT2LMHeadModel/GPT2Model[transformer]/NNCFEmbedding[wpe]/embedding_0"; "7 GPT2LMHeadModel/GPT2Model[transformer]/__add___0"; "8 GPT2LMHeadModel/GPT2Model[transformer]/Dropout[drop]/dropout_0"; "9 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_1]/layer_norm_0"; "10 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "11 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "12 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "13 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/split_0"; "14 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_0"; "15 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_0"; "16 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_1"; "17 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_1"; "18 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_2"; "19 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_2"; "20 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/transpose_0"; "21 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_0"; "22 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/full_0"; "23 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__truediv___0"; "24 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/full_1"; "25 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/where_0"; "26 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__add___0"; "27 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/softmax_0"; "28 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/type_0"; "29 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "30 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_1"; "31 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_3"; "32 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/contiguous_0"; "33 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_3"; "34 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "35 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "36 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "37 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "38 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___0"; "39 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_2]/layer_norm_0"; "40 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "41 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "42 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "43 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "44 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "45 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "46 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "47 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "48 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "49 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "50 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "51 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "52 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "53 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "54 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "55 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___1"; "56 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_1]/layer_norm_0"; "57 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "58 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "59 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "60 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/split_0"; "61 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_0"; "62 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_0"; "63 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_1"; "64 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_1"; "65 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_2"; "66 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_2"; "67 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/transpose_0"; "68 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_0"; "69 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/full_0"; "70 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__truediv___0"; "71 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/full_1"; "72 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/where_0"; "73 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__add___0"; "74 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/softmax_0"; "75 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/type_0"; "76 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "77 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_1"; "78 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_3"; "79 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/contiguous_0"; "80 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_3"; "81 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "82 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "83 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "84 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "85 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___0"; "86 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_2]/layer_norm_0"; "87 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "88 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "89 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "90 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "91 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "92 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "93 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "94 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "95 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "96 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "97 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "98 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "99 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "100 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "101 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "102 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___1"; "103 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_1]/layer_norm_0"; "104 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "105 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "106 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "107 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/split_0"; "108 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_0"; "109 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_0"; "110 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_1"; "111 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_1"; "112 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_2"; "113 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_2"; "114 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/transpose_0"; "115 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_0"; "116 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/full_0"; "117 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__truediv___0"; "118 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/full_1"; "119 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/where_0"; "120 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__add___0"; "121 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/softmax_0"; "122 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/type_0"; "123 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "124 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_1"; "125 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_3"; "126 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/contiguous_0"; "127 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_3"; "128 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "129 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "130 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "131 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "132 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___0"; "133 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_2]/layer_norm_0"; "134 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "135 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "136 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "137 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "138 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "139 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "140 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "141 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "142 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "143 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "144 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "145 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "146 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "147 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "148 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "149 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___1"; "150 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_1]/layer_norm_0"; "151 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "152 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "153 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "154 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/split_0"; "155 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_0"; "156 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_0"; "157 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_1"; "158 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_1"; "159 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_2"; "160 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_2"; "161 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/transpose_0"; "162 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_0"; "163 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/full_0"; "164 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__truediv___0"; "165 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/full_1"; "166 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/where_0"; "167 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__add___0"; "168 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/softmax_0"; "169 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/type_0"; "170 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "171 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_1"; "172 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_3"; "173 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/contiguous_0"; "174 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_3"; "175 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "176 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "177 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "178 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "179 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___0"; "180 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_2]/layer_norm_0"; "181 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "182 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "183 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "184 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "185 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "186 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "187 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "188 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "189 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "190 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "191 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "192 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "193 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "194 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "195 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "196 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___1"; "197 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_1]/layer_norm_0"; "198 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "199 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "200 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "201 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/split_0"; "202 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_0"; "203 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_0"; "204 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_1"; "205 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_1"; "206 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_2"; "207 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_2"; "208 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/transpose_0"; "209 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_0"; "210 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/full_0"; "211 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__truediv___0"; "212 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/full_1"; "213 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/where_0"; "214 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__add___0"; "215 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/softmax_0"; "216 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/type_0"; "217 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "218 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_1"; "219 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_3"; "220 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/contiguous_0"; "221 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_3"; "222 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "223 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "224 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "225 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "226 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___0"; "227 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_2]/layer_norm_0"; "228 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "229 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "230 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "231 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "232 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "233 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "234 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "235 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "236 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "237 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "238 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "239 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "240 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "241 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "242 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "243 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___1"; "244 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_1]/layer_norm_0"; "245 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "246 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "247 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "248 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/split_0"; "249 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_0"; "250 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_0"; "251 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_1"; "252 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_1"; "253 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_2"; "254 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_2"; "255 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/transpose_0"; "256 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_0"; "257 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/full_0"; "258 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__truediv___0"; "259 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/full_1"; "260 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/where_0"; "261 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__add___0"; "262 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/softmax_0"; "263 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/type_0"; "264 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "265 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_1"; "266 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_3"; "267 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/contiguous_0"; "268 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_3"; "269 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "270 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "271 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "272 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "273 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___0"; "274 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_2]/layer_norm_0"; "275 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "276 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "277 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "278 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "279 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "280 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "281 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "282 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "283 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "284 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "285 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "286 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "287 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "288 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "289 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "290 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___1"; "291 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_1]/layer_norm_0"; "292 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "293 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "294 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "295 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/split_0"; "296 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_0"; "297 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_0"; "298 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_1"; "299 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_1"; "300 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_2"; "301 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_2"; "302 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/transpose_0"; "303 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_0"; "304 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/full_0"; "305 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__truediv___0"; "306 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/full_1"; "307 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/where_0"; "308 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__add___0"; "309 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/softmax_0"; "310 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/type_0"; "311 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "312 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_1"; "313 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_3"; "314 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/contiguous_0"; "315 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_3"; "316 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "317 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "318 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "319 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "320 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___0"; "321 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_2]/layer_norm_0"; "322 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "323 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "324 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "325 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "326 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "327 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "328 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "329 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "330 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "331 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "332 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "333 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "334 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "335 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "336 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "337 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___1"; "338 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_1]/layer_norm_0"; "339 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "340 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "341 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "342 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/split_0"; "343 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_0"; "344 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_0"; "345 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_1"; "346 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_1"; "347 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_2"; "348 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_2"; "349 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/transpose_0"; "350 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_0"; "351 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/full_0"; "352 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__truediv___0"; "353 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/full_1"; "354 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/where_0"; "355 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__add___0"; "356 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/softmax_0"; "357 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/type_0"; "358 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "359 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_1"; "360 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_3"; "361 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/contiguous_0"; "362 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_3"; "363 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "364 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "365 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "366 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "367 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___0"; "368 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_2]/layer_norm_0"; "369 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "370 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "371 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "372 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "373 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "374 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "375 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "376 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "377 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "378 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "379 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "380 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "381 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "382 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "383 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "384 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___1"; "385 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_1]/layer_norm_0"; "386 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "387 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "388 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "389 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/split_0"; "390 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_0"; "391 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_0"; "392 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_1"; "393 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_1"; "394 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_2"; "395 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_2"; "396 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/transpose_0"; "397 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_0"; "398 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/full_0"; "399 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__truediv___0"; "400 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/full_1"; "401 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/where_0"; "402 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__add___0"; "403 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/softmax_0"; "404 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/type_0"; "405 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "406 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_1"; "407 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_3"; "408 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/contiguous_0"; "409 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_3"; "410 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "411 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "412 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "413 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "414 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___0"; "415 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_2]/layer_norm_0"; "416 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "417 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "418 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "419 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "420 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "421 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "422 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "423 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "424 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "425 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "426 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "427 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "428 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "429 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "430 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "431 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___1"; "432 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_1]/layer_norm_0"; "433 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "434 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "435 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "436 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/split_0"; "437 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_0"; "438 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_0"; "439 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_1"; "440 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_1"; "441 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_2"; "442 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_2"; "443 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/transpose_0"; "444 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_0"; "445 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/full_0"; "446 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__truediv___0"; "447 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/full_1"; "448 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/where_0"; "449 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__add___0"; "450 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/softmax_0"; "451 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/type_0"; "452 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "453 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_1"; "454 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_3"; "455 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/contiguous_0"; "456 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_3"; "457 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "458 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "459 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "460 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "461 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___0"; "462 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_2]/layer_norm_0"; "463 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "464 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "465 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "466 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "467 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "468 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "469 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "470 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "471 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "472 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "473 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "474 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "475 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "476 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "477 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "478 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___1"; "479 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_1]/layer_norm_0"; "480 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "481 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "482 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "483 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/split_0"; "484 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_0"; "485 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_0"; "486 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_1"; "487 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_1"; "488 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_2"; "489 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_2"; "490 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/transpose_0"; "491 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_0"; "492 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/full_0"; "493 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__truediv___0"; "494 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/full_1"; "495 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/where_0"; "496 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__add___0"; "497 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/softmax_0"; "498 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/type_0"; "499 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "500 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_1"; "501 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_3"; "502 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/contiguous_0"; "503 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_3"; "504 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "505 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "506 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "507 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "508 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___0"; "509 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_2]/layer_norm_0"; "510 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "511 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "512 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "513 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "514 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "515 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "516 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "517 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "518 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "519 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "520 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "521 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "522 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "523 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "524 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "525 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___1"; "526 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_1]/layer_norm_0"; "527 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_0"; "528 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0"; "529 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_1"; "530 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/split_0"; "531 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_0"; "532 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_0"; "533 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_1"; "534 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_1"; "535 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_2"; "536 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_2"; "537 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/transpose_0"; "538 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_0"; "539 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/full_0"; "540 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__truediv___0"; "541 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/full_1"; "542 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/where_0"; "543 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__add___0"; "544 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/softmax_0"; "545 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/type_0"; "546 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0"; "547 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_1"; "548 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_3"; "549 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/contiguous_0"; "550 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_3"; "551 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_0"; "552 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0"; "553 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_1"; "554 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0"; "555 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___0"; "556 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_2]/layer_norm_0"; "557 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0"; "558 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0"; "559 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1"; "560 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0"; "561 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0"; "562 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1"; "563 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0"; "564 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2"; "565 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0"; "566 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0"; "567 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0"; "568 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0"; "569 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0"; "570 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1"; "571 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0"; "572 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___1"; "573 GPT2LMHeadModel/GPT2Model[transformer]/NNCFLayerNorm[ln_f]/layer_norm_0"; "574 GPT2LMHeadModel/GPT2Model[transformer]/view_2"; "575 GPT2LMHeadModel/NNCFLinear[lm_head]/linear_0"; "576 /nncf_model_output_0"; "577 /nncf_model_output_1"; "578 /nncf_model_output_2"; "579 /nncf_model_output_3"; "580 /nncf_model_output_4"; "581 /nncf_model_output_5"; "582 /nncf_model_output_6"; "583 /nncf_model_output_7"; "584 /nncf_model_output_8"; "585 /nncf_model_output_9"; "586 /nncf_model_output_10"; "587 /nncf_model_output_11"; "588 /nncf_model_output_12"; "589 /nncf_model_output_13"; "590 /nncf_model_output_14"; "591 /nncf_model_output_15"; "592 /nncf_model_output_16"; "593 /nncf_model_output_17"; "594 /nncf_model_output_18"; "595 /nncf_model_output_19"; "596 /nncf_model_output_20"; "597 /nncf_model_output_21"; "598 /nncf_model_output_22"; "599 /nncf_model_output_23"; "600 /nncf_model_output_24"; "0 /nncf_model_input_0" -> "2 GPT2LMHeadModel/GPT2Model[transformer]/view_0" [label="(8, 1024) \n0 -> 0", style=dashed]; "1 /nncf_model_input_1" -> "3 GPT2LMHeadModel/GPT2Model[transformer]/view_1" [label="(8, 1024) \n0 -> 0", style=dashed]; "2 GPT2LMHeadModel/GPT2Model[transformer]/view_0" -> "5 GPT2LMHeadModel/GPT2Model[transformer]/NNCFEmbedding[wte]/embedding_0" [label="(8, 1024) \n0 -> 0", style=dashed]; "3 GPT2LMHeadModel/GPT2Model[transformer]/view_1" -> "4 GPT2LMHeadModel/GPT2Model[transformer]/__getitem___0" [label="(8, 1024) \n0 -> 0", style=dashed]; "5 GPT2LMHeadModel/GPT2Model[transformer]/NNCFEmbedding[wte]/embedding_0" -> "7 GPT2LMHeadModel/GPT2Model[transformer]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "6 GPT2LMHeadModel/GPT2Model[transformer]/NNCFEmbedding[wpe]/embedding_0" -> "7 GPT2LMHeadModel/GPT2Model[transformer]/__add___0" [label="(1, 1024, 768) \n0 -> 1", style=solid]; "7 GPT2LMHeadModel/GPT2Model[transformer]/__add___0" -> "8 GPT2LMHeadModel/GPT2Model[transformer]/Dropout[drop]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "8 GPT2LMHeadModel/GPT2Model[transformer]/Dropout[drop]/dropout_0" -> "9 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "8 GPT2LMHeadModel/GPT2Model[transformer]/Dropout[drop]/dropout_0" -> "38 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "9 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "10 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "10 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "11 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "11 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "12 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "12 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "13 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "13 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/split_0" -> "14 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "13 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/split_0" -> "16 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "13 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/split_0" -> "18 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "14 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_0" -> "15 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "15 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_0" -> "21 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "16 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_1" -> "17 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "17 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_1" -> "20 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "17 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_1" -> "576 /nncf_model_output_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "18 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_2" -> "19 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "19 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_2" -> "30 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "19 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_2" -> "577 /nncf_model_output_1" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "20 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/transpose_0" -> "21 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "21 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_0" -> "23 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "22 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/full_0" -> "23 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "23 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__truediv___0" -> "25 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "24 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/full_1" -> "25 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "25 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/where_0" -> "26 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "26 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/__add___0" -> "27 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "27 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/softmax_0" -> "28 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "28 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/type_0" -> "29 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "29 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "30 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "30 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/matmul_1" -> "31 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "31 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/permute_3" -> "32 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "32 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/contiguous_0" -> "33 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "33 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/view_3" -> "34 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "34 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "35 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "35 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "36 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "36 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "37 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "37 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "38 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "38 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___0" -> "39 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "38 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___0" -> "55 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "39 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "40 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "40 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "41 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "41 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "42 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "42 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "43 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "42 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "44 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "42 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "46 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "43 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "50 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "44 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "45 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "45 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "46 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "46 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "47 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "47 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "48 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "48 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "49 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "49 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "50 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "50 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "51 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "51 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "52 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "52 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "53 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "53 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "54 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "54 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "55 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "55 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___1" -> "56 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "55 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[0]/__add___1" -> "85 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "56 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "57 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "57 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "58 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "58 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "59 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "59 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "60 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "60 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/split_0" -> "61 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "60 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/split_0" -> "63 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "60 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/split_0" -> "65 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "61 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_0" -> "62 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "62 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_0" -> "68 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "63 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_1" -> "64 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "64 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_1" -> "67 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "64 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_1" -> "578 /nncf_model_output_2" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "65 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_2" -> "66 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "66 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_2" -> "77 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "66 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_2" -> "579 /nncf_model_output_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "67 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/transpose_0" -> "68 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "68 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_0" -> "70 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "69 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/full_0" -> "70 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "70 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__truediv___0" -> "72 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "71 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/full_1" -> "72 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "72 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/where_0" -> "73 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "73 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/__add___0" -> "74 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "74 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/softmax_0" -> "75 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "75 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/type_0" -> "76 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "76 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "77 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "77 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/matmul_1" -> "78 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "78 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/permute_3" -> "79 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "79 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/contiguous_0" -> "80 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "80 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/view_3" -> "81 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "81 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "82 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "82 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "83 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "83 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "84 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "84 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "85 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "85 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___0" -> "86 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "85 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___0" -> "102 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "86 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "87 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "87 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "88 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "88 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "89 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "89 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "90 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "89 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "91 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "89 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "93 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "90 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "97 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "91 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "92 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "92 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "93 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "93 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "94 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "94 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "95 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "95 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "96 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "96 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "97 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "97 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "98 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "98 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "99 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "99 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "100 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "100 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "101 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "101 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "102 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "102 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___1" -> "103 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "102 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[1]/__add___1" -> "132 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "103 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "104 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "104 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "105 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "105 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "106 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "106 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "107 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "107 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/split_0" -> "108 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "107 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/split_0" -> "110 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "107 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/split_0" -> "112 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "108 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_0" -> "109 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "109 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_0" -> "115 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "110 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_1" -> "111 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "111 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_1" -> "114 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "111 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_1" -> "580 /nncf_model_output_4" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "112 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_2" -> "113 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "113 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_2" -> "124 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "113 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_2" -> "581 /nncf_model_output_5" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "114 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/transpose_0" -> "115 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "115 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_0" -> "117 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "116 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/full_0" -> "117 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "117 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__truediv___0" -> "119 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "118 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/full_1" -> "119 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "119 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/where_0" -> "120 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "120 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/__add___0" -> "121 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "121 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/softmax_0" -> "122 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "122 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/type_0" -> "123 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "123 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "124 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "124 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/matmul_1" -> "125 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "125 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/permute_3" -> "126 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "126 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/contiguous_0" -> "127 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "127 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/view_3" -> "128 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "128 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "129 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "129 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "130 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "130 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "131 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "131 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "132 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "132 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___0" -> "133 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "132 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___0" -> "149 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "133 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "134 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "134 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "135 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "135 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "136 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "136 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "137 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "136 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "138 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "136 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "140 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "137 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "144 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "138 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "139 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "139 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "140 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "140 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "141 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "141 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "142 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "142 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "143 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "143 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "144 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "144 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "145 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "145 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "146 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "146 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "147 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "147 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "148 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "148 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "149 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "149 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___1" -> "150 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "149 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[2]/__add___1" -> "179 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "150 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "151 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "151 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "152 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "152 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "153 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "153 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "154 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "154 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/split_0" -> "155 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "154 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/split_0" -> "157 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "154 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/split_0" -> "159 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "155 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_0" -> "156 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "156 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_0" -> "162 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "157 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_1" -> "158 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "158 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_1" -> "161 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "158 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_1" -> "582 /nncf_model_output_6" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "159 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_2" -> "160 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "160 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_2" -> "171 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "160 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_2" -> "583 /nncf_model_output_7" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "161 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/transpose_0" -> "162 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "162 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_0" -> "164 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "163 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/full_0" -> "164 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "164 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__truediv___0" -> "166 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "165 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/full_1" -> "166 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "166 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/where_0" -> "167 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "167 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/__add___0" -> "168 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "168 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/softmax_0" -> "169 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "169 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/type_0" -> "170 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "170 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "171 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "171 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/matmul_1" -> "172 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "172 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/permute_3" -> "173 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "173 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/contiguous_0" -> "174 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "174 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/view_3" -> "175 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "175 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "176 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "176 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "177 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "177 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "178 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "178 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "179 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "179 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___0" -> "180 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "179 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___0" -> "196 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "180 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "181 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "181 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "182 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "182 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "183 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "183 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "184 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "183 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "185 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "183 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "187 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "184 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "191 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "185 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "186 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "186 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "187 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "187 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "188 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "188 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "189 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "189 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "190 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "190 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "191 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "191 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "192 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "192 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "193 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "193 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "194 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "194 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "195 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "195 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "196 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "196 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___1" -> "197 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "196 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[3]/__add___1" -> "226 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "197 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "198 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "198 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "199 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "199 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "200 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "200 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "201 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "201 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/split_0" -> "202 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "201 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/split_0" -> "204 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "201 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/split_0" -> "206 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "202 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_0" -> "203 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "203 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_0" -> "209 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "204 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_1" -> "205 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "205 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_1" -> "208 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "205 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_1" -> "584 /nncf_model_output_8" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "206 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_2" -> "207 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "207 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_2" -> "218 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "207 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_2" -> "585 /nncf_model_output_9" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "208 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/transpose_0" -> "209 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "209 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_0" -> "211 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "210 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/full_0" -> "211 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "211 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__truediv___0" -> "213 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "212 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/full_1" -> "213 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "213 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/where_0" -> "214 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "214 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/__add___0" -> "215 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "215 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/softmax_0" -> "216 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "216 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/type_0" -> "217 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "217 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "218 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "218 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/matmul_1" -> "219 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "219 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/permute_3" -> "220 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "220 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/contiguous_0" -> "221 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "221 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/view_3" -> "222 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "222 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "223 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "223 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "224 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "224 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "225 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "225 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "226 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "226 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___0" -> "227 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "226 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___0" -> "243 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "227 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "228 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "228 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "229 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "229 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "230 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "230 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "231 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "230 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "232 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "230 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "234 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "231 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "238 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "232 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "233 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "233 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "234 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "234 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "235 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "235 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "236 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "236 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "237 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "237 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "238 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "238 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "239 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "239 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "240 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "240 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "241 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "241 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "242 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "242 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "243 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "243 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___1" -> "244 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "243 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[4]/__add___1" -> "273 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "244 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "245 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "245 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "246 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "246 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "247 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "247 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "248 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "248 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/split_0" -> "249 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "248 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/split_0" -> "251 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "248 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/split_0" -> "253 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "249 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_0" -> "250 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "250 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_0" -> "256 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "251 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_1" -> "252 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "252 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_1" -> "255 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "252 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_1" -> "586 /nncf_model_output_10" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "253 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_2" -> "254 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "254 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_2" -> "265 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "254 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_2" -> "587 /nncf_model_output_11" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "255 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/transpose_0" -> "256 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "256 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_0" -> "258 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "257 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/full_0" -> "258 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "258 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__truediv___0" -> "260 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "259 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/full_1" -> "260 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "260 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/where_0" -> "261 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "261 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/__add___0" -> "262 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "262 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/softmax_0" -> "263 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "263 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/type_0" -> "264 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "264 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "265 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "265 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/matmul_1" -> "266 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "266 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/permute_3" -> "267 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "267 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/contiguous_0" -> "268 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "268 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/view_3" -> "269 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "269 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "270 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "270 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "271 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "271 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "272 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "272 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "273 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "273 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___0" -> "274 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "273 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___0" -> "290 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "274 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "275 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "275 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "276 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "276 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "277 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "277 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "278 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "277 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "279 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "277 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "281 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "278 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "285 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "279 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "280 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "280 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "281 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "281 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "282 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "282 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "283 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "283 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "284 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "284 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "285 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "285 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "286 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "286 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "287 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "287 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "288 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "288 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "289 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "289 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "290 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "290 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___1" -> "291 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "290 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[5]/__add___1" -> "320 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "291 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "292 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "292 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "293 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "293 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "294 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "294 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "295 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "295 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/split_0" -> "296 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "295 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/split_0" -> "298 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "295 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/split_0" -> "300 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "296 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_0" -> "297 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "297 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_0" -> "303 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "298 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_1" -> "299 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "299 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_1" -> "302 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "299 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_1" -> "588 /nncf_model_output_12" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "300 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_2" -> "301 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "301 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_2" -> "312 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "301 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_2" -> "589 /nncf_model_output_13" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "302 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/transpose_0" -> "303 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "303 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_0" -> "305 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "304 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/full_0" -> "305 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "305 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__truediv___0" -> "307 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "306 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/full_1" -> "307 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "307 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/where_0" -> "308 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "308 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/__add___0" -> "309 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "309 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/softmax_0" -> "310 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "310 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/type_0" -> "311 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "311 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "312 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "312 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/matmul_1" -> "313 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "313 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/permute_3" -> "314 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "314 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/contiguous_0" -> "315 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "315 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/view_3" -> "316 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "316 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "317 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "317 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "318 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "318 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "319 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "319 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "320 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "320 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___0" -> "321 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "320 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___0" -> "337 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "321 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "322 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "322 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "323 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "323 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "324 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "324 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "325 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "324 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "326 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "324 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "328 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "325 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "332 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "326 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "327 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "327 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "328 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "328 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "329 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "329 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "330 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "330 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "331 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "331 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "332 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "332 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "333 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "333 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "334 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "334 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "335 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "335 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "336 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "336 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "337 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "337 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___1" -> "338 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "337 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[6]/__add___1" -> "367 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "338 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "339 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "339 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "340 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "340 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "341 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "341 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "342 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "342 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/split_0" -> "343 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "342 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/split_0" -> "345 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "342 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/split_0" -> "347 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "343 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_0" -> "344 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "344 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_0" -> "350 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "345 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_1" -> "346 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "346 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_1" -> "349 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "346 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_1" -> "590 /nncf_model_output_14" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "347 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_2" -> "348 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "348 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_2" -> "359 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "348 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_2" -> "591 /nncf_model_output_15" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "349 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/transpose_0" -> "350 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "350 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_0" -> "352 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "351 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/full_0" -> "352 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "352 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__truediv___0" -> "354 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "353 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/full_1" -> "354 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "354 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/where_0" -> "355 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "355 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/__add___0" -> "356 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "356 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/softmax_0" -> "357 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "357 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/type_0" -> "358 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "358 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "359 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "359 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/matmul_1" -> "360 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "360 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/permute_3" -> "361 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "361 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/contiguous_0" -> "362 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "362 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/view_3" -> "363 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "363 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "364 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "364 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "365 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "365 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "366 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "366 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "367 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "367 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___0" -> "368 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "367 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___0" -> "384 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "368 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "369 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "369 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "370 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "370 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "371 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "371 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "372 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "371 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "373 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "371 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "375 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "372 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "379 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "373 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "374 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "374 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "375 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "375 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "376 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "376 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "377 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "377 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "378 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "378 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "379 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "379 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "380 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "380 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "381 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "381 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "382 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "382 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "383 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "383 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "384 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "384 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___1" -> "385 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "384 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[7]/__add___1" -> "414 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "385 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "386 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "386 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "387 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "387 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "388 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "388 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "389 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "389 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/split_0" -> "390 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "389 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/split_0" -> "392 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "389 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/split_0" -> "394 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "390 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_0" -> "391 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "391 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_0" -> "397 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "392 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_1" -> "393 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "393 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_1" -> "396 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "393 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_1" -> "592 /nncf_model_output_16" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "394 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_2" -> "395 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "395 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_2" -> "406 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "395 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_2" -> "593 /nncf_model_output_17" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "396 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/transpose_0" -> "397 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "397 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_0" -> "399 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "398 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/full_0" -> "399 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "399 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__truediv___0" -> "401 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "400 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/full_1" -> "401 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "401 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/where_0" -> "402 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "402 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/__add___0" -> "403 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "403 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/softmax_0" -> "404 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "404 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/type_0" -> "405 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "405 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "406 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "406 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/matmul_1" -> "407 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "407 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/permute_3" -> "408 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "408 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/contiguous_0" -> "409 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "409 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/view_3" -> "410 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "410 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "411 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "411 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "412 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "412 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "413 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "413 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "414 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "414 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___0" -> "415 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "414 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___0" -> "431 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "415 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "416 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "416 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "417 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "417 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "418 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "418 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "419 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "418 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "420 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "418 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "422 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "419 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "426 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "420 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "421 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "421 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "422 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "422 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "423 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "423 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "424 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "424 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "425 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "425 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "426 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "426 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "427 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "427 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "428 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "428 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "429 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "429 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "430 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "430 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "431 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "431 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___1" -> "432 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "431 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[8]/__add___1" -> "461 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "432 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "433 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "433 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "434 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "434 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "435 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "435 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "436 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "436 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/split_0" -> "437 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "436 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/split_0" -> "439 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "436 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/split_0" -> "441 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "437 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_0" -> "438 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "438 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_0" -> "444 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "439 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_1" -> "440 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "440 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_1" -> "443 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "440 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_1" -> "594 /nncf_model_output_18" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "441 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_2" -> "442 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "442 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_2" -> "453 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "442 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_2" -> "595 /nncf_model_output_19" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "443 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/transpose_0" -> "444 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "444 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_0" -> "446 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "445 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/full_0" -> "446 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "446 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__truediv___0" -> "448 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "447 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/full_1" -> "448 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "448 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/where_0" -> "449 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "449 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/__add___0" -> "450 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "450 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/softmax_0" -> "451 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "451 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/type_0" -> "452 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "452 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "453 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "453 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/matmul_1" -> "454 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "454 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/permute_3" -> "455 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "455 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/contiguous_0" -> "456 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "456 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/view_3" -> "457 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "457 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "458 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "458 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "459 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "459 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "460 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "460 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "461 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "461 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___0" -> "462 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "461 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___0" -> "478 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "462 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "463 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "463 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "464 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "464 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "465 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "465 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "466 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "465 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "467 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "465 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "469 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "466 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "473 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "467 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "468 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "468 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "469 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "469 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "470 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "470 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "471 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "471 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "472 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "472 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "473 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "473 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "474 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "474 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "475 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "475 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "476 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "476 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "477 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "477 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "478 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "478 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___1" -> "479 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "478 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[9]/__add___1" -> "508 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "479 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "480 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "480 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "481 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "481 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "482 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "482 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "483 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "483 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/split_0" -> "484 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "483 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/split_0" -> "486 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "483 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/split_0" -> "488 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "484 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_0" -> "485 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "485 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_0" -> "491 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "486 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_1" -> "487 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "487 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_1" -> "490 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "487 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_1" -> "596 /nncf_model_output_20" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "488 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_2" -> "489 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "489 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_2" -> "500 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "489 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_2" -> "597 /nncf_model_output_21" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "490 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/transpose_0" -> "491 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "491 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_0" -> "493 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "492 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/full_0" -> "493 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "493 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__truediv___0" -> "495 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "494 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/full_1" -> "495 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "495 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/where_0" -> "496 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "496 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/__add___0" -> "497 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "497 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/softmax_0" -> "498 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "498 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/type_0" -> "499 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "499 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "500 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "500 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/matmul_1" -> "501 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "501 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/permute_3" -> "502 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "502 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/contiguous_0" -> "503 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "503 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/view_3" -> "504 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "504 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "505 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "505 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "506 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "506 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "507 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "507 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "508 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "508 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___0" -> "509 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "508 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___0" -> "525 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "509 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "510 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "510 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "511 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "511 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "512 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "512 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "513 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "512 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "514 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "512 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "516 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "513 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "520 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "514 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "515 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "515 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "516 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "516 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "517 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "517 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "518 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "518 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "519 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "519 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "520 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "520 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "521 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "521 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "522 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "522 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "523 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "523 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "524 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "524 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "525 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "525 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___1" -> "526 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_1]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "525 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[10]/__add___1" -> "555 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___0" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "526 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_1]/layer_norm_0" -> "527 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "527 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_0" -> "528 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "528 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/addmm_0" -> "529 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" [label="(8192, 2304) \n0 -> 0", style=solid]; "529 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_attn]/view_1" -> "530 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/split_0" [label="(8, 1024, 2304) \n0 -> 0", style=solid]; "530 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/split_0" -> "531 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "530 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/split_0" -> "533 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_1" [label="(8, 1024, 768) \n1 -> 0", style=solid]; "530 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/split_0" -> "535 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_2" [label="(8, 1024, 768) \n2 -> 0", style=solid]; "531 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_0" -> "532 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "532 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_0" -> "538 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "533 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_1" -> "534 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_1" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "534 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_1" -> "537 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/transpose_0" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "534 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_1" -> "598 /nncf_model_output_22" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "535 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_2" -> "536 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_2" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "536 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_2" -> "547 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 64) \n0 -> 1", style=solid]; "536 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_2" -> "599 /nncf_model_output_23" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "537 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/transpose_0" -> "538 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_0" [label="(8, 12, 64, 1024) \n0 -> 1", style=solid]; "538 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_0" -> "540 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__truediv___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "539 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/full_0" -> "540 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__truediv___0" [label="() \n0 -> 1", style=solid]; "540 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__truediv___0" -> "542 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/where_0" [label="(8, 12, 1024, 1024) \n0 -> 1", style=solid]; "541 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/full_1" -> "542 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/where_0" [label="() \n0 -> 2", style=solid]; "542 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/where_0" -> "543 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__add___0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "543 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/__add___0" -> "544 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/softmax_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "544 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/softmax_0" -> "545 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/type_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "545 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/type_0" -> "546 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "546 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[attn_dropout]/dropout_0" -> "547 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_1" [label="(8, 12, 1024, 1024) \n0 -> 0", style=solid]; "547 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/matmul_1" -> "548 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_3" [label="(8, 12, 1024, 64) \n0 -> 0", style=solid]; "548 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/permute_3" -> "549 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/contiguous_0" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "549 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/contiguous_0" -> "550 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_3" [label="(8, 1024, 12, 64) \n0 -> 0", style=solid]; "550 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/view_3" -> "551 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "551 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_0" -> "552 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "552 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/addmm_0" -> "553 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "553 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Conv1D[c_proj]/view_1" -> "554 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "554 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2Attention[attn]/Dropout[resid_dropout]/dropout_0" -> "555 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "555 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___0" -> "556 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_2]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "555 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___0" -> "572 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___1" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "556 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/NNCFLayerNorm[ln_2]/layer_norm_0" -> "557 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "557 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_0" -> "558 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" [label="(8192, 768) \n0 -> 1", style=solid]; "558 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/addmm_0" -> "559 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" [label="(8192, 3072) \n0 -> 0", style=solid]; "559 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "560 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "559 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "561 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "559 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_fc]/view_1" -> "563 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "560 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___0" -> "567 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "561 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/pow_0" -> "562 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "562 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___1" -> "563 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "563 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__add___0" -> "564 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "564 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__rmul___2" -> "565 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "565 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/tanh_0" -> "566 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "566 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__radd___0" -> "567 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" [label="(8, 1024, 3072) \n0 -> 1", style=solid]; "567 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/NewGELUActivation[act]/__mul___0" -> "568 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" [label="(8, 1024, 3072) \n0 -> 0", style=solid]; "568 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_0" -> "569 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" [label="(8192, 3072) \n0 -> 1", style=solid]; "569 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/addmm_0" -> "570 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" [label="(8192, 768) \n0 -> 0", style=solid]; "570 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Conv1D[c_proj]/view_1" -> "571 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "571 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/GPT2MLP[mlp]/Dropout[dropout]/dropout_0" -> "572 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___1" [label="(8, 1024, 768) \n0 -> 1", style=solid]; "572 GPT2LMHeadModel/GPT2Model[transformer]/ModuleList[h]/GPT2Block[11]/__add___1" -> "573 GPT2LMHeadModel/GPT2Model[transformer]/NNCFLayerNorm[ln_f]/layer_norm_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "573 GPT2LMHeadModel/GPT2Model[transformer]/NNCFLayerNorm[ln_f]/layer_norm_0" -> "574 GPT2LMHeadModel/GPT2Model[transformer]/view_2" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "574 GPT2LMHeadModel/GPT2Model[transformer]/view_2" -> "575 GPT2LMHeadModel/NNCFLinear[lm_head]/linear_0" [label="(8, 1024, 768) \n0 -> 0", style=solid]; "575 GPT2LMHeadModel/NNCFLinear[lm_head]/linear_0" -> "600 /nncf_model_output_24" [label="(8, 1024, 50257) \n0 -> 0", style=solid]; }