bert_training.mlir

module attributes {torch.debug_module_name = "forward"} {
  func private @__torch__.torch.fx.graph_module.forward.__code_getter(%arg0: !torch.nn.Module<"__torch__.torch.fx.graph_module.forward">) -> !torch.str {
    %2 = torch.prim.GetAttr %arg0["_code"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.str
    return %2 : !torch.str
  }
  func private @__torch__.torch.fx.graph_module.forward.forward(%arg0: !torch.nn.Module<"__torch__.torch.fx.graph_module.forward">, %arg1: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg2: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg3: !torch.tensor {torch.type_bound = !torch.vtensor<[512,384],f32>}, %arg4: !torch.tensor {torch.type_bound = !torch.vtensor<[2,384],f32>}, %arg5: !torch.tensor {torch.type_bound = !torch.vtensor<[30522,384],f32>}, %arg6: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg7: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg8: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg9: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg10: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg11: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg12: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg13: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg14: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg15: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg16: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg17: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg18: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg19: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg20: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg21: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg22: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg23: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg24: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg25: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg26: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg27: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg28: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg29: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg30: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg31: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg32: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg33: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg34: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg35: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg36: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg37: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg38: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg39: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg40: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg41: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg42: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg43: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg44: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg45: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg46: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg47: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg48: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg49: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg50: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg51: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg52: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg53: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg54: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg55: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg56: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg57: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg58: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg59: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg60: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg61: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg62: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg63: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg64: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg65: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg66: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg67: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg68: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg69: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg70: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg71: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg72: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg73: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg74: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg75: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg76: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg77: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg78: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg79: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg80: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg81: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg82: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg83: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg84: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg85: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg86: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg87: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg88: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg89: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg90: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg91: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg92: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg93: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg94: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg95: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg96: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg97: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg98: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg99: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg100: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg101: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg102: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg103: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg104: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg105: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg106: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg107: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg108: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg109: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg110: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg111: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg112: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg113: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg114: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg115: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg116: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg117: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg118: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg119: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg120: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg121: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg122: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg123: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg124: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg125: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg126: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg127: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg128: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg129: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg130: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg131: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg132: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg133: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg134: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg135: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg136: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg137: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg138: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg139: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg140: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg141: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg142: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg143: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg144: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg145: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg146: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg147: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg148: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg149: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg150: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg151: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg152: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg153: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg154: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg155: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg156: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg157: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg158: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg159: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg160: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg161: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg162: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg163: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg164: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg165: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg166: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg167: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg168: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg169: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg170: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg171: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg172: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg173: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg174: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg175: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg176: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg177: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg178: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg179: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg180: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg181: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg182: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg183: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg184: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg185: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg186: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg187: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg188: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg189: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg190: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg191: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg192: !torch.tensor {torch.type_bound = !torch.vtensor<[1536],f32>}, %arg193: !torch.tensor {torch.type_bound = !torch.vtensor<[1536,384],f32>}, %arg194: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg195: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg196: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg197: !torch.tensor {torch.type_bound = !torch.vtensor<[384,1536],f32>}, %arg198: !torch.tensor {torch.type_bound = !torch.vtensor<[384],f32>}, %arg199: !torch.tensor {torch.type_bound = !torch.vtensor<[384,384],f32>}, %arg200: !torch.tensor {torch.type_bound = !torch.vtensor<[2],f32>}, %arg201: !torch.tensor {torch.type_bound = !torch.vtensor<[2,384],f32>}, %arg202: !torch.tensor {torch.type_bound = !torch.vtensor<[1,512],si64>}, %arg203: !torch.tensor {torch.type_bound = !torch.vtensor<[1,512],si64>}, %arg204: !torch.tensor {torch.type_bound = !torch.vtensor<[1,128],si64>}) -> !torch.tuple<tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor> {
    %true_0 = torch.constant.bool true
    %cpu = torch.constant.device "cpu"
    %none_1 = torch.constant.none
    %str_2 = torch.constant.str "none"
    %false = torch.constant.bool false
    %int0 = torch.constant.int 0
    %int9223372036854775807 = torch.constant.int 9223372036854775807
    %int1 = torch.constant.int 1
    %int128 = torch.constant.int 128
    %int384 = torch.constant.int 384
    %float9.999990e-13 = torch.constant.float 9.9999999999999998E-13
    %int12 = torch.constant.int 12
    %int32 = torch.constant.int 32
    %int2 = torch.constant.int 2
    %int3 = torch.constant.int 3
    %int-1 = torch.constant.int -1
    %int-2 = torch.constant.int -2
    %float5.656850e00 = torch.constant.float 5.6568542494923806
    %int16384 = torch.constant.int 16384
    %int1536 = torch.constant.int 1536
    %int6 = torch.constant.int 6
    %int196608 = torch.constant.int 196608
    %int49152 = torch.constant.int 49152
    %int4096 = torch.constant.int 4096
    %int512 = torch.constant.int 512
    %int30522 = torch.constant.int 30522
    %float-1.000000e-02 = torch.constant.float -1.000000e-02
    %2 = torch.aten.slice.Tensor %arg203, %int0, %int0, %int9223372036854775807, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int -> !torch.tensor
    %3 = torch.aten.slice.Tensor %2, %int1, %int0, %int128, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int -> !torch.tensor
    %4 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %5 = torch.aten.expand %3, %4, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %6 = torch.aten.slice.Tensor %arg202, %int0, %int0, %int9223372036854775807, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int -> !torch.tensor
    %7 = torch.aten.slice.Tensor %6, %int1, %int0, %int128, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int -> !torch.tensor
    %8 = torch.aten.embedding %arg5, %arg204, %int0, %false, %false : !torch.tensor, !torch.tensor, !torch.int, !torch.bool, !torch.bool -> !torch.tensor
    %9 = torch.aten.embedding %arg4, %5, %int-1, %false, %false : !torch.tensor, !torch.tensor, !torch.int, !torch.bool, !torch.bool -> !torch.tensor
    %10 = torch.aten.add.Tensor %8, %9, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %11 = torch.aten.embedding %arg3, %7, %int-1, %false, %false : !torch.tensor, !torch.tensor, !torch.int, !torch.bool, !torch.bool -> !torch.tensor
    %12 = torch.aten.add_.Tensor %10, %11, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %13 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0, %result1, %result2 = torch.aten.native_layer_norm %12, %13, %arg2, %arg1, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %14 = torch.aten.t %arg13 : !torch.tensor -> !torch.tensor
    %15 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %16 = torch.aten.view %result0, %15 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %17 = torch.aten.addmm %arg12, %16, %14, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %18 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %19 = torch.aten.view %17, %18 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %20 = torch.aten.t %arg11 : !torch.tensor -> !torch.tensor
    %21 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %22 = torch.aten.view %result0, %21 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %23 = torch.aten.addmm %arg10, %22, %20, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %24 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %25 = torch.aten.view %23, %24 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %26 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %27 = torch.aten.view %25, %26 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %28 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %29 = torch.aten.permute %27, %28 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %30 = torch.aten.t %arg15 : !torch.tensor -> !torch.tensor
    %31 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %32 = torch.aten.view %result0, %31 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %33 = torch.aten.addmm %arg14, %32, %30, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %34 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %35 = torch.aten.view %33, %34 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %36 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %37 = torch.aten.view %35, %36 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %38 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %39 = torch.aten.permute %37, %38 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %40 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %41 = torch.aten.view %19, %40 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %42 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %43 = torch.aten.permute %41, %42 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %44 = torch.aten.transpose.int %29, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %45 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %46 = torch.aten.expand %43, %45, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %47 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %48 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %49 = torch.aten._reshape_alias %46, %47, %48 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %50 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %51 = torch.aten.expand %44, %50, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %52 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %53 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %54 = torch.aten._reshape_alias %51, %52, %53 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %55 = torch.aten.bmm %49, %54 : !torch.tensor, !torch.tensor -> !torch.tensor
    %56 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %57 = torch.aten._unsafe_view %55, %56 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %58 = torch.aten.div.Scalar %57, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %59 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %60 = torch.aten.add.Tensor %58, %59, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %61 = torch.aten._softmax %60, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %62 = torch.aten.detach %61 : !torch.tensor -> !torch.tensor
    %63 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %64 = torch.aten.expand %61, %63, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %65 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %66 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %67 = torch.aten._reshape_alias %64, %65, %66 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %68 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %69 = torch.aten.expand %39, %68, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %70 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %71 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %72 = torch.aten._reshape_alias %69, %70, %71 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %73 = torch.aten.bmm %67, %72 : !torch.tensor, !torch.tensor -> !torch.tensor
    %74 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %75 = torch.aten._unsafe_view %73, %74 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %76 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %77 = torch.aten.permute %75, %76 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %78 = torch.aten.clone %77, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %79 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %80 = torch.aten.view %78, %79 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %81 = torch.aten.t %arg9 : !torch.tensor -> !torch.tensor
    %82 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %83 = torch.aten.view %80, %82 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %84 = torch.aten.addmm %arg8, %83, %81, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %85 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %86 = torch.aten.view %84, %85 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %87 = torch.aten.add.Tensor %86, %result0, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %88 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_3, %result1_4, %result2_5 = torch.aten.native_layer_norm %87, %88, %arg7, %arg6, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %89 = torch.aten.t %arg17 : !torch.tensor -> !torch.tensor
    %90 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %91 = torch.aten.view %result0_3, %90 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %92 = torch.aten.addmm %arg16, %91, %89, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %93 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %94 = torch.aten.view %92, %93 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %95 = torch.aten.gelu %94, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %96 = torch.aten.t %arg21 : !torch.tensor -> !torch.tensor
    %97 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %98 = torch.aten.view %95, %97 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %99 = torch.aten.addmm %arg20, %98, %96, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %100 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %101 = torch.aten.view %99, %100 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %102 = torch.aten.add.Tensor %101, %result0_3, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %103 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_6, %result1_7, %result2_8 = torch.aten.native_layer_norm %102, %103, %arg19, %arg18, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %104 = torch.aten.t %arg29 : !torch.tensor -> !torch.tensor
    %105 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %106 = torch.aten.view %result0_6, %105 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %107 = torch.aten.addmm %arg28, %106, %104, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %108 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %109 = torch.aten.view %107, %108 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %110 = torch.aten.t %arg27 : !torch.tensor -> !torch.tensor
    %111 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %112 = torch.aten.view %result0_6, %111 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %113 = torch.aten.addmm %arg26, %112, %110, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %114 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %115 = torch.aten.view %113, %114 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %116 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %117 = torch.aten.view %115, %116 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %118 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %119 = torch.aten.permute %117, %118 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %120 = torch.aten.t %arg31 : !torch.tensor -> !torch.tensor
    %121 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %122 = torch.aten.view %result0_6, %121 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %123 = torch.aten.addmm %arg30, %122, %120, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %124 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %125 = torch.aten.view %123, %124 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %126 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %127 = torch.aten.view %125, %126 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %128 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %129 = torch.aten.permute %127, %128 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %130 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %131 = torch.aten.view %109, %130 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %132 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %133 = torch.aten.permute %131, %132 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %134 = torch.aten.transpose.int %119, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %135 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %136 = torch.aten.expand %133, %135, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %137 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %138 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %139 = torch.aten._reshape_alias %136, %137, %138 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %140 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %141 = torch.aten.expand %134, %140, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %142 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %143 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %144 = torch.aten._reshape_alias %141, %142, %143 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %145 = torch.aten.bmm %139, %144 : !torch.tensor, !torch.tensor -> !torch.tensor
    %146 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %147 = torch.aten._unsafe_view %145, %146 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %148 = torch.aten.div.Scalar %147, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %149 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %150 = torch.aten.add.Tensor %148, %149, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %151 = torch.aten._softmax %150, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %152 = torch.aten.detach %151 : !torch.tensor -> !torch.tensor
    %153 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %154 = torch.aten.expand %151, %153, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %155 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %156 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %157 = torch.aten._reshape_alias %154, %155, %156 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %158 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %159 = torch.aten.expand %129, %158, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %160 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %161 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %162 = torch.aten._reshape_alias %159, %160, %161 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %163 = torch.aten.bmm %157, %162 : !torch.tensor, !torch.tensor -> !torch.tensor
    %164 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %165 = torch.aten._unsafe_view %163, %164 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %166 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %167 = torch.aten.permute %165, %166 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %168 = torch.aten.clone %167, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %169 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %170 = torch.aten.view %168, %169 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %171 = torch.aten.t %arg25 : !torch.tensor -> !torch.tensor
    %172 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %173 = torch.aten.view %170, %172 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %174 = torch.aten.addmm %arg24, %173, %171, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %175 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %176 = torch.aten.view %174, %175 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %177 = torch.aten.add.Tensor %176, %result0_6, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %178 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_9, %result1_10, %result2_11 = torch.aten.native_layer_norm %177, %178, %arg23, %arg22, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %179 = torch.aten.t %arg33 : !torch.tensor -> !torch.tensor
    %180 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %181 = torch.aten.view %result0_9, %180 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %182 = torch.aten.addmm %arg32, %181, %179, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %183 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %184 = torch.aten.view %182, %183 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %185 = torch.aten.gelu %184, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %186 = torch.aten.t %arg37 : !torch.tensor -> !torch.tensor
    %187 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %188 = torch.aten.view %185, %187 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %189 = torch.aten.addmm %arg36, %188, %186, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %190 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %191 = torch.aten.view %189, %190 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %192 = torch.aten.add.Tensor %191, %result0_9, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %193 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_12, %result1_13, %result2_14 = torch.aten.native_layer_norm %192, %193, %arg35, %arg34, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %194 = torch.aten.t %arg77 : !torch.tensor -> !torch.tensor
    %195 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %196 = torch.aten.view %result0_12, %195 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %197 = torch.aten.addmm %arg76, %196, %194, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %198 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %199 = torch.aten.view %197, %198 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %200 = torch.aten.t %arg75 : !torch.tensor -> !torch.tensor
    %201 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %202 = torch.aten.view %result0_12, %201 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %203 = torch.aten.addmm %arg74, %202, %200, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %204 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %205 = torch.aten.view %203, %204 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %206 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %207 = torch.aten.view %205, %206 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %208 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %209 = torch.aten.permute %207, %208 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %210 = torch.aten.t %arg79 : !torch.tensor -> !torch.tensor
    %211 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %212 = torch.aten.view %result0_12, %211 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %213 = torch.aten.addmm %arg78, %212, %210, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %214 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %215 = torch.aten.view %213, %214 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %216 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %217 = torch.aten.view %215, %216 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %218 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %219 = torch.aten.permute %217, %218 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %220 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %221 = torch.aten.view %199, %220 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %222 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %223 = torch.aten.permute %221, %222 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %224 = torch.aten.transpose.int %209, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %225 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %226 = torch.aten.expand %223, %225, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %227 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %228 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %229 = torch.aten._reshape_alias %226, %227, %228 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %230 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %231 = torch.aten.expand %224, %230, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %232 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %233 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %234 = torch.aten._reshape_alias %231, %232, %233 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %235 = torch.aten.bmm %229, %234 : !torch.tensor, !torch.tensor -> !torch.tensor
    %236 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %237 = torch.aten._unsafe_view %235, %236 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %238 = torch.aten.div.Scalar %237, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %239 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %240 = torch.aten.add.Tensor %238, %239, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %241 = torch.aten._softmax %240, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %242 = torch.aten.detach %241 : !torch.tensor -> !torch.tensor
    %243 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %244 = torch.aten.expand %241, %243, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %245 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %246 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %247 = torch.aten._reshape_alias %244, %245, %246 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %248 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %249 = torch.aten.expand %219, %248, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %250 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %251 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %252 = torch.aten._reshape_alias %249, %250, %251 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %253 = torch.aten.bmm %247, %252 : !torch.tensor, !torch.tensor -> !torch.tensor
    %254 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %255 = torch.aten._unsafe_view %253, %254 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %256 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %257 = torch.aten.permute %255, %256 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %258 = torch.aten.clone %257, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %259 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %260 = torch.aten.view %258, %259 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %261 = torch.aten.t %arg73 : !torch.tensor -> !torch.tensor
    %262 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %263 = torch.aten.view %260, %262 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %264 = torch.aten.addmm %arg72, %263, %261, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %265 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %266 = torch.aten.view %264, %265 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %267 = torch.aten.add.Tensor %266, %result0_12, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %268 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_15, %result1_16, %result2_17 = torch.aten.native_layer_norm %267, %268, %arg71, %arg70, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %269 = torch.aten.t %arg81 : !torch.tensor -> !torch.tensor
    %270 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %271 = torch.aten.view %result0_15, %270 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %272 = torch.aten.addmm %arg80, %271, %269, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %273 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %274 = torch.aten.view %272, %273 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %275 = torch.aten.gelu %274, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %276 = torch.aten.t %arg85 : !torch.tensor -> !torch.tensor
    %277 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %278 = torch.aten.view %275, %277 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %279 = torch.aten.addmm %arg84, %278, %276, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %280 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %281 = torch.aten.view %279, %280 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %282 = torch.aten.add.Tensor %281, %result0_15, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %283 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_18, %result1_19, %result2_20 = torch.aten.native_layer_norm %282, %283, %arg83, %arg82, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %284 = torch.aten.t %arg93 : !torch.tensor -> !torch.tensor
    %285 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %286 = torch.aten.view %result0_18, %285 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %287 = torch.aten.addmm %arg92, %286, %284, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %288 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %289 = torch.aten.view %287, %288 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %290 = torch.aten.t %arg91 : !torch.tensor -> !torch.tensor
    %291 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %292 = torch.aten.view %result0_18, %291 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %293 = torch.aten.addmm %arg90, %292, %290, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %294 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %295 = torch.aten.view %293, %294 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %296 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %297 = torch.aten.view %295, %296 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %298 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %299 = torch.aten.permute %297, %298 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %300 = torch.aten.t %arg95 : !torch.tensor -> !torch.tensor
    %301 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %302 = torch.aten.view %result0_18, %301 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %303 = torch.aten.addmm %arg94, %302, %300, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %304 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %305 = torch.aten.view %303, %304 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %306 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %307 = torch.aten.view %305, %306 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %308 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %309 = torch.aten.permute %307, %308 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %310 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %311 = torch.aten.view %289, %310 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %312 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %313 = torch.aten.permute %311, %312 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %314 = torch.aten.transpose.int %299, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %315 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %316 = torch.aten.expand %313, %315, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %317 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %318 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %319 = torch.aten._reshape_alias %316, %317, %318 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %320 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %321 = torch.aten.expand %314, %320, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %322 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %323 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %324 = torch.aten._reshape_alias %321, %322, %323 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %325 = torch.aten.bmm %319, %324 : !torch.tensor, !torch.tensor -> !torch.tensor
    %326 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %327 = torch.aten._unsafe_view %325, %326 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %328 = torch.aten.div.Scalar %327, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %329 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %330 = torch.aten.add.Tensor %328, %329, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %331 = torch.aten._softmax %330, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %332 = torch.aten.detach %331 : !torch.tensor -> !torch.tensor
    %333 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %334 = torch.aten.expand %331, %333, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %335 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %336 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %337 = torch.aten._reshape_alias %334, %335, %336 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %338 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %339 = torch.aten.expand %309, %338, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %340 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %341 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %342 = torch.aten._reshape_alias %339, %340, %341 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %343 = torch.aten.bmm %337, %342 : !torch.tensor, !torch.tensor -> !torch.tensor
    %344 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %345 = torch.aten._unsafe_view %343, %344 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %346 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %347 = torch.aten.permute %345, %346 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %348 = torch.aten.clone %347, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %349 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %350 = torch.aten.view %348, %349 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %351 = torch.aten.t %arg89 : !torch.tensor -> !torch.tensor
    %352 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %353 = torch.aten.view %350, %352 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %354 = torch.aten.addmm %arg88, %353, %351, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %355 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %356 = torch.aten.view %354, %355 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %357 = torch.aten.add.Tensor %356, %result0_18, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %358 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_21, %result1_22, %result2_23 = torch.aten.native_layer_norm %357, %358, %arg87, %arg86, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %359 = torch.aten.t %arg97 : !torch.tensor -> !torch.tensor
    %360 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %361 = torch.aten.view %result0_21, %360 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %362 = torch.aten.addmm %arg96, %361, %359, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %363 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %364 = torch.aten.view %362, %363 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %365 = torch.aten.gelu %364, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %366 = torch.aten.t %arg101 : !torch.tensor -> !torch.tensor
    %367 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %368 = torch.aten.view %365, %367 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %369 = torch.aten.addmm %arg100, %368, %366, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %370 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %371 = torch.aten.view %369, %370 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %372 = torch.aten.add.Tensor %371, %result0_21, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %373 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_24, %result1_25, %result2_26 = torch.aten.native_layer_norm %372, %373, %arg99, %arg98, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %374 = torch.aten.t %arg109 : !torch.tensor -> !torch.tensor
    %375 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %376 = torch.aten.view %result0_24, %375 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %377 = torch.aten.addmm %arg108, %376, %374, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %378 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %379 = torch.aten.view %377, %378 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %380 = torch.aten.t %arg107 : !torch.tensor -> !torch.tensor
    %381 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %382 = torch.aten.view %result0_24, %381 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %383 = torch.aten.addmm %arg106, %382, %380, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %384 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %385 = torch.aten.view %383, %384 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %386 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %387 = torch.aten.view %385, %386 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %388 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %389 = torch.aten.permute %387, %388 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %390 = torch.aten.t %arg111 : !torch.tensor -> !torch.tensor
    %391 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %392 = torch.aten.view %result0_24, %391 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %393 = torch.aten.addmm %arg110, %392, %390, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %394 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %395 = torch.aten.view %393, %394 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %396 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %397 = torch.aten.view %395, %396 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %398 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %399 = torch.aten.permute %397, %398 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %400 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %401 = torch.aten.view %379, %400 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %402 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %403 = torch.aten.permute %401, %402 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %404 = torch.aten.transpose.int %389, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %405 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %406 = torch.aten.expand %403, %405, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %407 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %408 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %409 = torch.aten._reshape_alias %406, %407, %408 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %410 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %411 = torch.aten.expand %404, %410, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %412 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %413 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %414 = torch.aten._reshape_alias %411, %412, %413 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %415 = torch.aten.bmm %409, %414 : !torch.tensor, !torch.tensor -> !torch.tensor
    %416 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %417 = torch.aten._unsafe_view %415, %416 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %418 = torch.aten.div.Scalar %417, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %419 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %420 = torch.aten.add.Tensor %418, %419, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %421 = torch.aten._softmax %420, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %422 = torch.aten.detach %421 : !torch.tensor -> !torch.tensor
    %423 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %424 = torch.aten.expand %421, %423, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %425 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %426 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %427 = torch.aten._reshape_alias %424, %425, %426 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %428 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %429 = torch.aten.expand %399, %428, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %430 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %431 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %432 = torch.aten._reshape_alias %429, %430, %431 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %433 = torch.aten.bmm %427, %432 : !torch.tensor, !torch.tensor -> !torch.tensor
    %434 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %435 = torch.aten._unsafe_view %433, %434 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %436 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %437 = torch.aten.permute %435, %436 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %438 = torch.aten.clone %437, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %439 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %440 = torch.aten.view %438, %439 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %441 = torch.aten.t %arg105 : !torch.tensor -> !torch.tensor
    %442 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %443 = torch.aten.view %440, %442 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %444 = torch.aten.addmm %arg104, %443, %441, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %445 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %446 = torch.aten.view %444, %445 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %447 = torch.aten.add.Tensor %446, %result0_24, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %448 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_27, %result1_28, %result2_29 = torch.aten.native_layer_norm %447, %448, %arg103, %arg102, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %449 = torch.aten.t %arg113 : !torch.tensor -> !torch.tensor
    %450 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %451 = torch.aten.view %result0_27, %450 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %452 = torch.aten.addmm %arg112, %451, %449, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %453 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %454 = torch.aten.view %452, %453 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %455 = torch.aten.gelu %454, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %456 = torch.aten.t %arg117 : !torch.tensor -> !torch.tensor
    %457 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %458 = torch.aten.view %455, %457 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %459 = torch.aten.addmm %arg116, %458, %456, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %460 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %461 = torch.aten.view %459, %460 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %462 = torch.aten.add.Tensor %461, %result0_27, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %463 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_30, %result1_31, %result2_32 = torch.aten.native_layer_norm %462, %463, %arg115, %arg114, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %464 = torch.aten.t %arg125 : !torch.tensor -> !torch.tensor
    %465 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %466 = torch.aten.view %result0_30, %465 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %467 = torch.aten.addmm %arg124, %466, %464, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %468 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %469 = torch.aten.view %467, %468 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %470 = torch.aten.t %arg123 : !torch.tensor -> !torch.tensor
    %471 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %472 = torch.aten.view %result0_30, %471 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %473 = torch.aten.addmm %arg122, %472, %470, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %474 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %475 = torch.aten.view %473, %474 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %476 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %477 = torch.aten.view %475, %476 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %478 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %479 = torch.aten.permute %477, %478 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %480 = torch.aten.t %arg127 : !torch.tensor -> !torch.tensor
    %481 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %482 = torch.aten.view %result0_30, %481 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %483 = torch.aten.addmm %arg126, %482, %480, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %484 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %485 = torch.aten.view %483, %484 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %486 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %487 = torch.aten.view %485, %486 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %488 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %489 = torch.aten.permute %487, %488 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %490 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %491 = torch.aten.view %469, %490 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %492 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %493 = torch.aten.permute %491, %492 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %494 = torch.aten.transpose.int %479, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %495 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %496 = torch.aten.expand %493, %495, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %497 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %498 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %499 = torch.aten._reshape_alias %496, %497, %498 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %500 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %501 = torch.aten.expand %494, %500, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %502 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %503 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %504 = torch.aten._reshape_alias %501, %502, %503 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %505 = torch.aten.bmm %499, %504 : !torch.tensor, !torch.tensor -> !torch.tensor
    %506 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %507 = torch.aten._unsafe_view %505, %506 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %508 = torch.aten.div.Scalar %507, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %509 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %510 = torch.aten.add.Tensor %508, %509, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %511 = torch.aten._softmax %510, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %512 = torch.aten.detach %511 : !torch.tensor -> !torch.tensor
    %513 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %514 = torch.aten.expand %511, %513, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %515 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %516 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %517 = torch.aten._reshape_alias %514, %515, %516 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %518 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %519 = torch.aten.expand %489, %518, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %520 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %521 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %522 = torch.aten._reshape_alias %519, %520, %521 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %523 = torch.aten.bmm %517, %522 : !torch.tensor, !torch.tensor -> !torch.tensor
    %524 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %525 = torch.aten._unsafe_view %523, %524 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %526 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %527 = torch.aten.permute %525, %526 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %528 = torch.aten.clone %527, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %529 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %530 = torch.aten.view %528, %529 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %531 = torch.aten.t %arg121 : !torch.tensor -> !torch.tensor
    %532 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %533 = torch.aten.view %530, %532 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %534 = torch.aten.addmm %arg120, %533, %531, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %535 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %536 = torch.aten.view %534, %535 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %537 = torch.aten.add.Tensor %536, %result0_30, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %538 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_33, %result1_34, %result2_35 = torch.aten.native_layer_norm %537, %538, %arg119, %arg118, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %539 = torch.aten.t %arg129 : !torch.tensor -> !torch.tensor
    %540 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %541 = torch.aten.view %result0_33, %540 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %542 = torch.aten.addmm %arg128, %541, %539, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %543 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %544 = torch.aten.view %542, %543 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %545 = torch.aten.gelu %544, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %546 = torch.aten.t %arg133 : !torch.tensor -> !torch.tensor
    %547 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %548 = torch.aten.view %545, %547 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %549 = torch.aten.addmm %arg132, %548, %546, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %550 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %551 = torch.aten.view %549, %550 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %552 = torch.aten.add.Tensor %551, %result0_33, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %553 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_36, %result1_37, %result2_38 = torch.aten.native_layer_norm %552, %553, %arg131, %arg130, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %554 = torch.aten.t %arg141 : !torch.tensor -> !torch.tensor
    %555 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %556 = torch.aten.view %result0_36, %555 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %557 = torch.aten.addmm %arg140, %556, %554, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %558 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %559 = torch.aten.view %557, %558 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %560 = torch.aten.t %arg139 : !torch.tensor -> !torch.tensor
    %561 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %562 = torch.aten.view %result0_36, %561 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %563 = torch.aten.addmm %arg138, %562, %560, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %564 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %565 = torch.aten.view %563, %564 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %566 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %567 = torch.aten.view %565, %566 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %568 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %569 = torch.aten.permute %567, %568 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %570 = torch.aten.t %arg143 : !torch.tensor -> !torch.tensor
    %571 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %572 = torch.aten.view %result0_36, %571 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %573 = torch.aten.addmm %arg142, %572, %570, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %574 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %575 = torch.aten.view %573, %574 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %576 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %577 = torch.aten.view %575, %576 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %578 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %579 = torch.aten.permute %577, %578 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %580 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %581 = torch.aten.view %559, %580 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %582 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %583 = torch.aten.permute %581, %582 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %584 = torch.aten.transpose.int %569, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %585 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %586 = torch.aten.expand %583, %585, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %587 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %588 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %589 = torch.aten._reshape_alias %586, %587, %588 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %590 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %591 = torch.aten.expand %584, %590, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %592 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %593 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %594 = torch.aten._reshape_alias %591, %592, %593 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %595 = torch.aten.bmm %589, %594 : !torch.tensor, !torch.tensor -> !torch.tensor
    %596 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %597 = torch.aten._unsafe_view %595, %596 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %598 = torch.aten.div.Scalar %597, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %599 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %600 = torch.aten.add.Tensor %598, %599, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %601 = torch.aten._softmax %600, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %602 = torch.aten.detach %601 : !torch.tensor -> !torch.tensor
    %603 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %604 = torch.aten.expand %601, %603, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %605 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %606 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %607 = torch.aten._reshape_alias %604, %605, %606 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %608 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %609 = torch.aten.expand %579, %608, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %610 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %611 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %612 = torch.aten._reshape_alias %609, %610, %611 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %613 = torch.aten.bmm %607, %612 : !torch.tensor, !torch.tensor -> !torch.tensor
    %614 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %615 = torch.aten._unsafe_view %613, %614 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %616 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %617 = torch.aten.permute %615, %616 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %618 = torch.aten.clone %617, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %619 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %620 = torch.aten.view %618, %619 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %621 = torch.aten.t %arg137 : !torch.tensor -> !torch.tensor
    %622 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %623 = torch.aten.view %620, %622 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %624 = torch.aten.addmm %arg136, %623, %621, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %625 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %626 = torch.aten.view %624, %625 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %627 = torch.aten.add.Tensor %626, %result0_36, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %628 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_39, %result1_40, %result2_41 = torch.aten.native_layer_norm %627, %628, %arg135, %arg134, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %629 = torch.aten.t %arg145 : !torch.tensor -> !torch.tensor
    %630 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %631 = torch.aten.view %result0_39, %630 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %632 = torch.aten.addmm %arg144, %631, %629, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %633 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %634 = torch.aten.view %632, %633 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %635 = torch.aten.gelu %634, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %636 = torch.aten.t %arg149 : !torch.tensor -> !torch.tensor
    %637 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %638 = torch.aten.view %635, %637 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %639 = torch.aten.addmm %arg148, %638, %636, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %640 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %641 = torch.aten.view %639, %640 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %642 = torch.aten.add.Tensor %641, %result0_39, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %643 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_42, %result1_43, %result2_44 = torch.aten.native_layer_norm %642, %643, %arg147, %arg146, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %644 = torch.aten.t %arg157 : !torch.tensor -> !torch.tensor
    %645 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %646 = torch.aten.view %result0_42, %645 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %647 = torch.aten.addmm %arg156, %646, %644, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %648 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %649 = torch.aten.view %647, %648 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %650 = torch.aten.t %arg155 : !torch.tensor -> !torch.tensor
    %651 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %652 = torch.aten.view %result0_42, %651 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %653 = torch.aten.addmm %arg154, %652, %650, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %654 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %655 = torch.aten.view %653, %654 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %656 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %657 = torch.aten.view %655, %656 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %658 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %659 = torch.aten.permute %657, %658 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %660 = torch.aten.t %arg159 : !torch.tensor -> !torch.tensor
    %661 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %662 = torch.aten.view %result0_42, %661 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %663 = torch.aten.addmm %arg158, %662, %660, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %664 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %665 = torch.aten.view %663, %664 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %666 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %667 = torch.aten.view %665, %666 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %668 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %669 = torch.aten.permute %667, %668 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %670 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %671 = torch.aten.view %649, %670 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %672 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %673 = torch.aten.permute %671, %672 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %674 = torch.aten.transpose.int %659, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %675 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %676 = torch.aten.expand %673, %675, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %677 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %678 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %679 = torch.aten._reshape_alias %676, %677, %678 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %680 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %681 = torch.aten.expand %674, %680, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %682 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %683 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %684 = torch.aten._reshape_alias %681, %682, %683 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %685 = torch.aten.bmm %679, %684 : !torch.tensor, !torch.tensor -> !torch.tensor
    %686 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %687 = torch.aten._unsafe_view %685, %686 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %688 = torch.aten.div.Scalar %687, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %689 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %690 = torch.aten.add.Tensor %688, %689, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %691 = torch.aten._softmax %690, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %692 = torch.aten.detach %691 : !torch.tensor -> !torch.tensor
    %693 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %694 = torch.aten.expand %691, %693, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %695 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %696 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %697 = torch.aten._reshape_alias %694, %695, %696 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %698 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %699 = torch.aten.expand %669, %698, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %700 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %701 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %702 = torch.aten._reshape_alias %699, %700, %701 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %703 = torch.aten.bmm %697, %702 : !torch.tensor, !torch.tensor -> !torch.tensor
    %704 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %705 = torch.aten._unsafe_view %703, %704 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %706 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %707 = torch.aten.permute %705, %706 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %708 = torch.aten.clone %707, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %709 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %710 = torch.aten.view %708, %709 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %711 = torch.aten.t %arg153 : !torch.tensor -> !torch.tensor
    %712 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %713 = torch.aten.view %710, %712 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %714 = torch.aten.addmm %arg152, %713, %711, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %715 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %716 = torch.aten.view %714, %715 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %717 = torch.aten.add.Tensor %716, %result0_42, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %718 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_45, %result1_46, %result2_47 = torch.aten.native_layer_norm %717, %718, %arg151, %arg150, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %719 = torch.aten.t %arg161 : !torch.tensor -> !torch.tensor
    %720 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %721 = torch.aten.view %result0_45, %720 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %722 = torch.aten.addmm %arg160, %721, %719, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %723 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %724 = torch.aten.view %722, %723 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %725 = torch.aten.gelu %724, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %726 = torch.aten.t %arg165 : !torch.tensor -> !torch.tensor
    %727 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %728 = torch.aten.view %725, %727 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %729 = torch.aten.addmm %arg164, %728, %726, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %730 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %731 = torch.aten.view %729, %730 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %732 = torch.aten.add.Tensor %731, %result0_45, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %733 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_48, %result1_49, %result2_50 = torch.aten.native_layer_norm %732, %733, %arg163, %arg162, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %734 = torch.aten.t %arg173 : !torch.tensor -> !torch.tensor
    %735 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %736 = torch.aten.view %result0_48, %735 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %737 = torch.aten.addmm %arg172, %736, %734, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %738 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %739 = torch.aten.view %737, %738 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %740 = torch.aten.t %arg171 : !torch.tensor -> !torch.tensor
    %741 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %742 = torch.aten.view %result0_48, %741 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %743 = torch.aten.addmm %arg170, %742, %740, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %744 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %745 = torch.aten.view %743, %744 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %746 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %747 = torch.aten.view %745, %746 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %748 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %749 = torch.aten.permute %747, %748 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %750 = torch.aten.t %arg175 : !torch.tensor -> !torch.tensor
    %751 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %752 = torch.aten.view %result0_48, %751 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %753 = torch.aten.addmm %arg174, %752, %750, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %754 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %755 = torch.aten.view %753, %754 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %756 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %757 = torch.aten.view %755, %756 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %758 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %759 = torch.aten.permute %757, %758 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %760 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %761 = torch.aten.view %739, %760 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %762 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %763 = torch.aten.permute %761, %762 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %764 = torch.aten.transpose.int %749, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %765 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %766 = torch.aten.expand %763, %765, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %767 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %768 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %769 = torch.aten._reshape_alias %766, %767, %768 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %770 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %771 = torch.aten.expand %764, %770, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %772 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %773 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %774 = torch.aten._reshape_alias %771, %772, %773 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %775 = torch.aten.bmm %769, %774 : !torch.tensor, !torch.tensor -> !torch.tensor
    %776 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %777 = torch.aten._unsafe_view %775, %776 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %778 = torch.aten.div.Scalar %777, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %779 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %780 = torch.aten.add.Tensor %778, %779, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %781 = torch.aten._softmax %780, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %782 = torch.aten.detach %781 : !torch.tensor -> !torch.tensor
    %783 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %784 = torch.aten.expand %781, %783, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %785 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %786 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %787 = torch.aten._reshape_alias %784, %785, %786 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %788 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %789 = torch.aten.expand %759, %788, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %790 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %791 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %792 = torch.aten._reshape_alias %789, %790, %791 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %793 = torch.aten.bmm %787, %792 : !torch.tensor, !torch.tensor -> !torch.tensor
    %794 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %795 = torch.aten._unsafe_view %793, %794 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %796 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %797 = torch.aten.permute %795, %796 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %798 = torch.aten.clone %797, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %799 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %800 = torch.aten.view %798, %799 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %801 = torch.aten.t %arg169 : !torch.tensor -> !torch.tensor
    %802 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %803 = torch.aten.view %800, %802 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %804 = torch.aten.addmm %arg168, %803, %801, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %805 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %806 = torch.aten.view %804, %805 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %807 = torch.aten.add.Tensor %806, %result0_48, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %808 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_51, %result1_52, %result2_53 = torch.aten.native_layer_norm %807, %808, %arg167, %arg166, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %809 = torch.aten.t %arg177 : !torch.tensor -> !torch.tensor
    %810 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %811 = torch.aten.view %result0_51, %810 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %812 = torch.aten.addmm %arg176, %811, %809, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %813 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %814 = torch.aten.view %812, %813 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %815 = torch.aten.gelu %814, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %816 = torch.aten.t %arg181 : !torch.tensor -> !torch.tensor
    %817 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %818 = torch.aten.view %815, %817 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %819 = torch.aten.addmm %arg180, %818, %816, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %820 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %821 = torch.aten.view %819, %820 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %822 = torch.aten.add.Tensor %821, %result0_51, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %823 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_54, %result1_55, %result2_56 = torch.aten.native_layer_norm %822, %823, %arg179, %arg178, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %824 = torch.aten.t %arg189 : !torch.tensor -> !torch.tensor
    %825 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %826 = torch.aten.view %result0_54, %825 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %827 = torch.aten.addmm %arg188, %826, %824, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %828 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %829 = torch.aten.view %827, %828 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %830 = torch.aten.t %arg187 : !torch.tensor -> !torch.tensor
    %831 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %832 = torch.aten.view %result0_54, %831 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %833 = torch.aten.addmm %arg186, %832, %830, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %834 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %835 = torch.aten.view %833, %834 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %836 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %837 = torch.aten.view %835, %836 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %838 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %839 = torch.aten.permute %837, %838 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %840 = torch.aten.t %arg191 : !torch.tensor -> !torch.tensor
    %841 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %842 = torch.aten.view %result0_54, %841 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %843 = torch.aten.addmm %arg190, %842, %840, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %844 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %845 = torch.aten.view %843, %844 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %846 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %847 = torch.aten.view %845, %846 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %848 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %849 = torch.aten.permute %847, %848 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %850 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %851 = torch.aten.view %829, %850 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %852 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %853 = torch.aten.permute %851, %852 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %854 = torch.aten.transpose.int %839, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %855 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %856 = torch.aten.expand %853, %855, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %857 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %858 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %859 = torch.aten._reshape_alias %856, %857, %858 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %860 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %861 = torch.aten.expand %854, %860, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %862 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %863 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %864 = torch.aten._reshape_alias %861, %862, %863 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %865 = torch.aten.bmm %859, %864 : !torch.tensor, !torch.tensor -> !torch.tensor
    %866 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %867 = torch.aten._unsafe_view %865, %866 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %868 = torch.aten.div.Scalar %867, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %869 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %870 = torch.aten.add.Tensor %868, %869, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %871 = torch.aten._softmax %870, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %872 = torch.aten.detach %871 : !torch.tensor -> !torch.tensor
    %873 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %874 = torch.aten.expand %871, %873, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %875 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %876 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %877 = torch.aten._reshape_alias %874, %875, %876 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %878 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %879 = torch.aten.expand %849, %878, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %880 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %881 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %882 = torch.aten._reshape_alias %879, %880, %881 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %883 = torch.aten.bmm %877, %882 : !torch.tensor, !torch.tensor -> !torch.tensor
    %884 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %885 = torch.aten._unsafe_view %883, %884 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %886 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %887 = torch.aten.permute %885, %886 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %888 = torch.aten.clone %887, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %889 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %890 = torch.aten.view %888, %889 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %891 = torch.aten.t %arg185 : !torch.tensor -> !torch.tensor
    %892 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %893 = torch.aten.view %890, %892 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %894 = torch.aten.addmm %arg184, %893, %891, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %895 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %896 = torch.aten.view %894, %895 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %897 = torch.aten.add.Tensor %896, %result0_54, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %898 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_57, %result1_58, %result2_59 = torch.aten.native_layer_norm %897, %898, %arg183, %arg182, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %899 = torch.aten.t %arg193 : !torch.tensor -> !torch.tensor
    %900 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %901 = torch.aten.view %result0_57, %900 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %902 = torch.aten.addmm %arg192, %901, %899, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %903 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %904 = torch.aten.view %902, %903 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %905 = torch.aten.gelu %904, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %906 = torch.aten.t %arg197 : !torch.tensor -> !torch.tensor
    %907 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %908 = torch.aten.view %905, %907 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %909 = torch.aten.addmm %arg196, %908, %906, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %910 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %911 = torch.aten.view %909, %910 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %912 = torch.aten.add.Tensor %911, %result0_57, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %913 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_60, %result1_61, %result2_62 = torch.aten.native_layer_norm %912, %913, %arg195, %arg194, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %914 = torch.aten.t %arg45 : !torch.tensor -> !torch.tensor
    %915 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %916 = torch.aten.view %result0_60, %915 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %917 = torch.aten.addmm %arg44, %916, %914, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %918 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %919 = torch.aten.view %917, %918 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %920 = torch.aten.t %arg43 : !torch.tensor -> !torch.tensor
    %921 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %922 = torch.aten.view %result0_60, %921 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %923 = torch.aten.addmm %arg42, %922, %920, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %924 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %925 = torch.aten.view %923, %924 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %926 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %927 = torch.aten.view %925, %926 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %928 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %929 = torch.aten.permute %927, %928 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %930 = torch.aten.t %arg47 : !torch.tensor -> !torch.tensor
    %931 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %932 = torch.aten.view %result0_60, %931 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %933 = torch.aten.addmm %arg46, %932, %930, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %934 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %935 = torch.aten.view %933, %934 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %936 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %937 = torch.aten.view %935, %936 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %938 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %939 = torch.aten.permute %937, %938 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %940 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %941 = torch.aten.view %919, %940 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %942 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %943 = torch.aten.permute %941, %942 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %944 = torch.aten.transpose.int %929, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %945 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %946 = torch.aten.expand %943, %945, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %947 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %948 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %949 = torch.aten._reshape_alias %946, %947, %948 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %950 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %951 = torch.aten.expand %944, %950, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %952 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %953 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %954 = torch.aten._reshape_alias %951, %952, %953 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %955 = torch.aten.bmm %949, %954 : !torch.tensor, !torch.tensor -> !torch.tensor
    %956 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %957 = torch.aten._unsafe_view %955, %956 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %958 = torch.aten.div.Scalar %957, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %959 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %960 = torch.aten.add.Tensor %958, %959, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %961 = torch.aten._softmax %960, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %962 = torch.aten.detach %961 : !torch.tensor -> !torch.tensor
    %963 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %964 = torch.aten.expand %961, %963, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %965 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %966 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %967 = torch.aten._reshape_alias %964, %965, %966 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %968 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %969 = torch.aten.expand %939, %968, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %970 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %971 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %972 = torch.aten._reshape_alias %969, %970, %971 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %973 = torch.aten.bmm %967, %972 : !torch.tensor, !torch.tensor -> !torch.tensor
    %974 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %975 = torch.aten._unsafe_view %973, %974 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %976 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %977 = torch.aten.permute %975, %976 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %978 = torch.aten.clone %977, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %979 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %980 = torch.aten.view %978, %979 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %981 = torch.aten.t %arg41 : !torch.tensor -> !torch.tensor
    %982 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %983 = torch.aten.view %980, %982 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %984 = torch.aten.addmm %arg40, %983, %981, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %985 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %986 = torch.aten.view %984, %985 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %987 = torch.aten.add.Tensor %986, %result0_60, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %988 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_63, %result1_64, %result2_65 = torch.aten.native_layer_norm %987, %988, %arg39, %arg38, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %989 = torch.aten.t %arg49 : !torch.tensor -> !torch.tensor
    %990 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %991 = torch.aten.view %result0_63, %990 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %992 = torch.aten.addmm %arg48, %991, %989, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %993 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %994 = torch.aten.view %992, %993 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %995 = torch.aten.gelu %994, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %996 = torch.aten.t %arg53 : !torch.tensor -> !torch.tensor
    %997 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %998 = torch.aten.view %995, %997 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %999 = torch.aten.addmm %arg52, %998, %996, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1000 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1001 = torch.aten.view %999, %1000 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1002 = torch.aten.add.Tensor %1001, %result0_63, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1003 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_66, %result1_67, %result2_68 = torch.aten.native_layer_norm %1002, %1003, %arg51, %arg50, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %1004 = torch.aten.t %arg61 : !torch.tensor -> !torch.tensor
    %1005 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1006 = torch.aten.view %result0_66, %1005 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1007 = torch.aten.addmm %arg60, %1006, %1004, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1008 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1009 = torch.aten.view %1007, %1008 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1010 = torch.aten.t %arg59 : !torch.tensor -> !torch.tensor
    %1011 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1012 = torch.aten.view %result0_66, %1011 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1013 = torch.aten.addmm %arg58, %1012, %1010, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1014 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1015 = torch.aten.view %1013, %1014 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1016 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1017 = torch.aten.view %1015, %1016 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1018 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1019 = torch.aten.permute %1017, %1018 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1020 = torch.aten.t %arg63 : !torch.tensor -> !torch.tensor
    %1021 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1022 = torch.aten.view %result0_66, %1021 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1023 = torch.aten.addmm %arg62, %1022, %1020, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1024 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1025 = torch.aten.view %1023, %1024 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1026 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1027 = torch.aten.view %1025, %1026 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1028 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1029 = torch.aten.permute %1027, %1028 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1030 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1031 = torch.aten.view %1009, %1030 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1032 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1033 = torch.aten.permute %1031, %1032 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1034 = torch.aten.transpose.int %1019, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1035 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1036 = torch.aten.expand %1033, %1035, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %1037 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1038 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1039 = torch.aten._reshape_alias %1036, %1037, %1038 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1040 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1041 = torch.aten.expand %1034, %1040, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %1042 = torch.prim.ListConstruct %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1043 = torch.prim.ListConstruct %int32, %int1, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1044 = torch.aten._reshape_alias %1041, %1042, %1043 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1045 = torch.aten.bmm %1039, %1044 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1046 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1047 = torch.aten._unsafe_view %1045, %1046 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1048 = torch.aten.div.Scalar %1047, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %1049 = torch.prim.GetAttr %arg0["_tensor_constant0"] : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward"> -> !torch.tensor
    %1050 = torch.aten.add.Tensor %1048, %1049, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1051 = torch.aten._softmax %1050, %int-1, %false : !torch.tensor, !torch.int, !torch.bool -> !torch.tensor
    %1052 = torch.aten.detach %1051 : !torch.tensor -> !torch.tensor
    %1053 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1054 = torch.aten.expand %1051, %1053, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %1055 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1056 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1057 = torch.aten._reshape_alias %1054, %1055, %1056 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1058 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1059 = torch.aten.expand %1029, %1058, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %1060 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1061 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1062 = torch.aten._reshape_alias %1059, %1060, %1061 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1063 = torch.aten.bmm %1057, %1062 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1064 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1065 = torch.aten._unsafe_view %1063, %1064 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1066 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1067 = torch.aten.permute %1065, %1066 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1068 = torch.aten.clone %1067, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1069 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1070 = torch.aten.view %1068, %1069 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1071 = torch.aten.t %arg57 : !torch.tensor -> !torch.tensor
    %1072 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1073 = torch.aten.view %1070, %1072 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1074 = torch.aten.addmm %arg56, %1073, %1071, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1075 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1076 = torch.aten.view %1074, %1075 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1077 = torch.aten.add.Tensor %1076, %result0_66, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1078 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_69, %result1_70, %result2_71 = torch.aten.native_layer_norm %1077, %1078, %arg55, %arg54, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %1079 = torch.aten.t %arg65 : !torch.tensor -> !torch.tensor
    %1080 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1081 = torch.aten.view %result0_69, %1080 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1082 = torch.aten.addmm %arg64, %1081, %1079, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1083 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1084 = torch.aten.view %1082, %1083 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1085 = torch.aten.gelu %1084, %str_2 : !torch.tensor, !torch.str -> !torch.tensor
    %1086 = torch.aten.t %arg69 : !torch.tensor -> !torch.tensor
    %1087 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %1088 = torch.aten.view %1085, %1087 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1089 = torch.aten.addmm %arg68, %1088, %1086, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1090 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1091 = torch.aten.view %1089, %1090 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1092 = torch.aten.add.Tensor %1091, %result0_69, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1093 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %result0_72, %result1_73, %result2_74 = torch.aten.native_layer_norm %1092, %1093, %arg67, %arg66, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %1094 = torch.aten.slice.Tensor %result0_72, %int0, %int0, %int9223372036854775807, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int -> !torch.tensor
    %1095 = torch.aten.select.int %1094, %int1, %int0 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1096 = torch.aten.t %arg199 : !torch.tensor -> !torch.tensor
    %1097 = torch.aten.addmm %arg198, %1095, %1096, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1098 = torch.aten.tanh %1097 : !torch.tensor -> !torch.tensor
    %1099 = torch.aten.detach %1098 : !torch.tensor -> !torch.tensor
    %1100 = torch.aten.t %arg201 : !torch.tensor -> !torch.tensor
    %1101 = torch.aten.addmm %arg200, %1098, %1100, %int1, %int1 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1102 = torch.aten.sum %1101, %none_1 : !torch.tensor, !torch.none -> !torch.tensor
    %1103 = torch.aten.ones_like %1102, %int6, %int0, %cpu, %false, %int1 : !torch.tensor, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.int -> !torch.tensor
    %1104 = torch.prim.ListConstruct %int1, %int2 : (!torch.int, !torch.int) -> !torch.list<int>
    %1105 = torch.aten.expand %1103, %1104, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %1106 = torch.aten.t %1100 : !torch.tensor -> !torch.tensor
    %1107 = torch.aten.mm %1105, %1106 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1108 = torch.aten.t %1105 : !torch.tensor -> !torch.tensor
    %1109 = torch.aten.mm %1108, %1098 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1110 = torch.aten.t %1109 : !torch.tensor -> !torch.tensor
    %1111 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1112 = torch.aten.sum.dim_IntList %1105, %1111, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1113 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1114 = torch.aten.view %1112, %1113 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1115 = torch.aten.detach %1114 : !torch.tensor -> !torch.tensor
    %1116 = torch.aten.detach %1115 : !torch.tensor -> !torch.tensor
    %1117 = torch.aten.t %1110 : !torch.tensor -> !torch.tensor
    %1118 = torch.aten.detach %1117 : !torch.tensor -> !torch.tensor
    %1119 = torch.aten.detach %1118 : !torch.tensor -> !torch.tensor
    %1120 = torch.aten.detach %1099 : !torch.tensor -> !torch.tensor
    %1121 = torch.aten.tanh_backward %1107, %1120 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1122 = torch.aten.t %1096 : !torch.tensor -> !torch.tensor
    %1123 = torch.aten.mm %1121, %1122 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1124 = torch.aten.t %1121 : !torch.tensor -> !torch.tensor
    %1125 = torch.aten.mm %1124, %1095 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1126 = torch.aten.t %1125 : !torch.tensor -> !torch.tensor
    %1127 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1128 = torch.aten.sum.dim_IntList %1121, %1127, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1129 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1130 = torch.aten.view %1128, %1129 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1131 = torch.aten.detach %1130 : !torch.tensor -> !torch.tensor
    %1132 = torch.aten.detach %1131 : !torch.tensor -> !torch.tensor
    %1133 = torch.aten.t %1126 : !torch.tensor -> !torch.tensor
    %1134 = torch.aten.detach %1133 : !torch.tensor -> !torch.tensor
    %1135 = torch.aten.detach %1134 : !torch.tensor -> !torch.tensor
    %1136 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1137 = torch.aten.new_zeros %1123, %1136, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %1138 = torch.operator "aten.select_scatter"(%1137, %1123, %int1, %int0) : (!torch.tensor, !torch.tensor, !torch.int, !torch.int) -> !torch.tensor
    %1139 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1140 = torch.aten.new_zeros %1138, %1139, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %1141 = torch.operator "aten.slice_scatter"(%1140, %1138, %int0, %int0, %int9223372036854775807, %int1) : (!torch.tensor, !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.int) -> !torch.tensor
    %1142 = torch.aten.sub.Tensor %1092, %result1_73, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1143 = torch.aten.mul.Tensor %1142, %result2_74 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1144 = torch.aten.mul.Tensor %1141, %arg67 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1145 = torch.aten.mul.Scalar %1144, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1146 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1147 = torch.aten.sum.dim_IntList %1144, %1146, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1148 = torch.aten.mul.Tensor %1144, %1143 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1149 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1150 = torch.aten.sum.dim_IntList %1148, %1149, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1151 = torch.aten.mul.Tensor %1143, %1150 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1152 = torch.aten.sub.Tensor %1145, %1147, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1153 = torch.aten.sub.Tensor %1152, %1151, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1154 = torch.aten.div.Scalar %result2_74, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1155 = torch.aten.mul.Tensor %1154, %1153 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1156 = torch.aten.mul.Tensor %1141, %1143 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1157 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1158 = torch.aten.sum.dim_IntList %1156, %1157, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1159 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1160 = torch.aten.sum.dim_IntList %1141, %1159, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1161 = torch.aten.detach %1158 : !torch.tensor -> !torch.tensor
    %1162 = torch.aten.detach %1161 : !torch.tensor -> !torch.tensor
    %1163 = torch.aten.detach %1160 : !torch.tensor -> !torch.tensor
    %1164 = torch.aten.detach %1163 : !torch.tensor -> !torch.tensor
    %1165 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1166 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1167 = torch.aten._reshape_alias %1155, %1165, %1166 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1168 = torch.aten.t %1086 : !torch.tensor -> !torch.tensor
    %1169 = torch.aten.mm %1167, %1168 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1170 = torch.aten.t %1167 : !torch.tensor -> !torch.tensor
    %1171 = torch.aten.mm %1170, %1088 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1172 = torch.aten.t %1171 : !torch.tensor -> !torch.tensor
    %1173 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1174 = torch.aten.sum.dim_IntList %1167, %1173, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1175 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1176 = torch.aten.view %1174, %1175 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1177 = torch.aten.detach %1176 : !torch.tensor -> !torch.tensor
    %1178 = torch.aten.detach %1177 : !torch.tensor -> !torch.tensor
    %1179 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1180 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1181 = torch.aten._reshape_alias %1169, %1179, %1180 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1182 = torch.aten.t %1172 : !torch.tensor -> !torch.tensor
    %1183 = torch.aten.detach %1182 : !torch.tensor -> !torch.tensor
    %1184 = torch.aten.detach %1183 : !torch.tensor -> !torch.tensor
    %1185 = torch.aten.gelu_backward %1181, %1084, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1186 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %1187 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1188 = torch.aten._reshape_alias %1185, %1186, %1187 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1189 = torch.aten.t %1079 : !torch.tensor -> !torch.tensor
    %1190 = torch.aten.mm %1188, %1189 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1191 = torch.aten.t %1188 : !torch.tensor -> !torch.tensor
    %1192 = torch.aten.mm %1191, %1081 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1193 = torch.aten.t %1192 : !torch.tensor -> !torch.tensor
    %1194 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1195 = torch.aten.sum.dim_IntList %1188, %1194, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1196 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %1197 = torch.aten.view %1195, %1196 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1198 = torch.aten.detach %1197 : !torch.tensor -> !torch.tensor
    %1199 = torch.aten.detach %1198 : !torch.tensor -> !torch.tensor
    %1200 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1201 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1202 = torch.aten._reshape_alias %1190, %1200, %1201 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1203 = torch.aten.add.Tensor %1155, %1202, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1204 = torch.aten.t %1193 : !torch.tensor -> !torch.tensor
    %1205 = torch.aten.detach %1204 : !torch.tensor -> !torch.tensor
    %1206 = torch.aten.detach %1205 : !torch.tensor -> !torch.tensor
    %1207 = torch.aten.sub.Tensor %1077, %result1_70, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1208 = torch.aten.mul.Tensor %1207, %result2_71 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1209 = torch.aten.mul.Tensor %1203, %arg55 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1210 = torch.aten.mul.Scalar %1209, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1211 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1212 = torch.aten.sum.dim_IntList %1209, %1211, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1213 = torch.aten.mul.Tensor %1209, %1208 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1214 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1215 = torch.aten.sum.dim_IntList %1213, %1214, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1216 = torch.aten.mul.Tensor %1208, %1215 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1217 = torch.aten.sub.Tensor %1210, %1212, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1218 = torch.aten.sub.Tensor %1217, %1216, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1219 = torch.aten.div.Scalar %result2_71, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1220 = torch.aten.mul.Tensor %1219, %1218 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1221 = torch.aten.mul.Tensor %1203, %1208 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1222 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1223 = torch.aten.sum.dim_IntList %1221, %1222, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1224 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1225 = torch.aten.sum.dim_IntList %1203, %1224, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1226 = torch.aten.detach %1223 : !torch.tensor -> !torch.tensor
    %1227 = torch.aten.detach %1226 : !torch.tensor -> !torch.tensor
    %1228 = torch.aten.detach %1225 : !torch.tensor -> !torch.tensor
    %1229 = torch.aten.detach %1228 : !torch.tensor -> !torch.tensor
    %1230 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1231 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1232 = torch.aten._reshape_alias %1220, %1230, %1231 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1233 = torch.aten.t %1071 : !torch.tensor -> !torch.tensor
    %1234 = torch.aten.mm %1232, %1233 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1235 = torch.aten.t %1232 : !torch.tensor -> !torch.tensor
    %1236 = torch.aten.mm %1235, %1073 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1237 = torch.aten.t %1236 : !torch.tensor -> !torch.tensor
    %1238 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1239 = torch.aten.sum.dim_IntList %1232, %1238, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1240 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1241 = torch.aten.view %1239, %1240 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1242 = torch.aten.detach %1241 : !torch.tensor -> !torch.tensor
    %1243 = torch.aten.detach %1242 : !torch.tensor -> !torch.tensor
    %1244 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1245 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1246 = torch.aten._reshape_alias %1234, %1244, %1245 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1247 = torch.aten.t %1237 : !torch.tensor -> !torch.tensor
    %1248 = torch.aten.detach %1247 : !torch.tensor -> !torch.tensor
    %1249 = torch.aten.detach %1248 : !torch.tensor -> !torch.tensor
    %1250 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1251 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1252 = torch.aten._reshape_alias %1246, %1250, %1251 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1253 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1254 = torch.aten.permute %1252, %1253 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1255 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1256 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1257 = torch.aten._reshape_alias %1254, %1255, %1256 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1258 = torch.aten.transpose.int %1057, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1259 = torch.aten.bmm %1258, %1257 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1260 = torch.aten.transpose.int %1062, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1261 = torch.aten.bmm %1257, %1260 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1262 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1263 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1264 = torch.aten._reshape_alias %1259, %1262, %1263 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1265 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1266 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1267 = torch.aten._reshape_alias %1261, %1265, %1266 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1268 = torch.aten.detach %1052 : !torch.tensor -> !torch.tensor
    %1269 = torch.aten._softmax_backward_data %1267, %1268, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1270 = torch.aten.div.Scalar %1269, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %1271 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1272 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1273 = torch.aten._reshape_alias %1270, %1271, %1272 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1274 = torch.aten.transpose.int %1039, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1275 = torch.aten.bmm %1274, %1273 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1276 = torch.aten.transpose.int %1044, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1277 = torch.aten.bmm %1273, %1276 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1278 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1279 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1280 = torch.aten._reshape_alias %1275, %1278, %1279 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1281 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1282 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1283 = torch.aten._reshape_alias %1277, %1281, %1282 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1284 = torch.aten.transpose.int %1280, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1285 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1286 = torch.aten.permute %1283, %1285 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1287 = torch.aten.clone %1286, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1288 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1289 = torch.aten._unsafe_view %1287, %1288 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1290 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1291 = torch.aten.permute %1264, %1290 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1292 = torch.aten.clone %1291, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1293 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1294 = torch.aten._unsafe_view %1292, %1293 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1295 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1296 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1297 = torch.aten._reshape_alias %1294, %1295, %1296 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1298 = torch.aten.t %1020 : !torch.tensor -> !torch.tensor
    %1299 = torch.aten.mm %1297, %1298 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1300 = torch.aten.t %1297 : !torch.tensor -> !torch.tensor
    %1301 = torch.aten.mm %1300, %1022 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1302 = torch.aten.t %1301 : !torch.tensor -> !torch.tensor
    %1303 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1304 = torch.aten.sum.dim_IntList %1297, %1303, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1305 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1306 = torch.aten.view %1304, %1305 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1307 = torch.aten.detach %1306 : !torch.tensor -> !torch.tensor
    %1308 = torch.aten.detach %1307 : !torch.tensor -> !torch.tensor
    %1309 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1310 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1311 = torch.aten._reshape_alias %1299, %1309, %1310 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1312 = torch.aten.add.Tensor %1220, %1311, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1313 = torch.aten.t %1302 : !torch.tensor -> !torch.tensor
    %1314 = torch.aten.detach %1313 : !torch.tensor -> !torch.tensor
    %1315 = torch.aten.detach %1314 : !torch.tensor -> !torch.tensor
    %1316 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1317 = torch.aten.permute %1284, %1316 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1318 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1319 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1320 = torch.aten._reshape_alias %1317, %1318, %1319 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1321 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1322 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %1323 = torch.aten._reshape_alias %1320, %1321, %1322 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1324 = torch.aten.t %1010 : !torch.tensor -> !torch.tensor
    %1325 = torch.aten.mm %1323, %1324 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1326 = torch.aten.t %1323 : !torch.tensor -> !torch.tensor
    %1327 = torch.aten.mm %1326, %1012 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1328 = torch.aten.t %1327 : !torch.tensor -> !torch.tensor
    %1329 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1330 = torch.aten.sum.dim_IntList %1323, %1329, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1331 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1332 = torch.aten.view %1330, %1331 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1333 = torch.aten.detach %1332 : !torch.tensor -> !torch.tensor
    %1334 = torch.aten.detach %1333 : !torch.tensor -> !torch.tensor
    %1335 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1336 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1337 = torch.aten._reshape_alias %1325, %1335, %1336 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1338 = torch.aten.add.Tensor %1312, %1337, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1339 = torch.aten.t %1328 : !torch.tensor -> !torch.tensor
    %1340 = torch.aten.detach %1339 : !torch.tensor -> !torch.tensor
    %1341 = torch.aten.detach %1340 : !torch.tensor -> !torch.tensor
    %1342 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1343 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1344 = torch.aten._reshape_alias %1289, %1342, %1343 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1345 = torch.aten.t %1004 : !torch.tensor -> !torch.tensor
    %1346 = torch.aten.mm %1344, %1345 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1347 = torch.aten.t %1344 : !torch.tensor -> !torch.tensor
    %1348 = torch.aten.mm %1347, %1006 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1349 = torch.aten.t %1348 : !torch.tensor -> !torch.tensor
    %1350 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1351 = torch.aten.sum.dim_IntList %1344, %1350, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1352 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1353 = torch.aten.view %1351, %1352 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1354 = torch.aten.detach %1353 : !torch.tensor -> !torch.tensor
    %1355 = torch.aten.detach %1354 : !torch.tensor -> !torch.tensor
    %1356 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1357 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1358 = torch.aten._reshape_alias %1346, %1356, %1357 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1359 = torch.aten.add.Tensor %1338, %1358, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1360 = torch.aten.t %1349 : !torch.tensor -> !torch.tensor
    %1361 = torch.aten.detach %1360 : !torch.tensor -> !torch.tensor
    %1362 = torch.aten.detach %1361 : !torch.tensor -> !torch.tensor
    %1363 = torch.aten.sub.Tensor %1002, %result1_67, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1364 = torch.aten.mul.Tensor %1363, %result2_68 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1365 = torch.aten.mul.Tensor %1359, %arg51 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1366 = torch.aten.mul.Scalar %1365, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1367 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1368 = torch.aten.sum.dim_IntList %1365, %1367, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1369 = torch.aten.mul.Tensor %1365, %1364 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1370 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1371 = torch.aten.sum.dim_IntList %1369, %1370, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1372 = torch.aten.mul.Tensor %1364, %1371 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1373 = torch.aten.sub.Tensor %1366, %1368, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1374 = torch.aten.sub.Tensor %1373, %1372, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1375 = torch.aten.div.Scalar %result2_68, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1376 = torch.aten.mul.Tensor %1375, %1374 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1377 = torch.aten.mul.Tensor %1359, %1364 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1378 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1379 = torch.aten.sum.dim_IntList %1377, %1378, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1380 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1381 = torch.aten.sum.dim_IntList %1359, %1380, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1382 = torch.aten.detach %1379 : !torch.tensor -> !torch.tensor
    %1383 = torch.aten.detach %1382 : !torch.tensor -> !torch.tensor
    %1384 = torch.aten.detach %1381 : !torch.tensor -> !torch.tensor
    %1385 = torch.aten.detach %1384 : !torch.tensor -> !torch.tensor
    %1386 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1387 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1388 = torch.aten._reshape_alias %1376, %1386, %1387 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1389 = torch.aten.t %996 : !torch.tensor -> !torch.tensor
    %1390 = torch.aten.mm %1388, %1389 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1391 = torch.aten.t %1388 : !torch.tensor -> !torch.tensor
    %1392 = torch.aten.mm %1391, %998 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1393 = torch.aten.t %1392 : !torch.tensor -> !torch.tensor
    %1394 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1395 = torch.aten.sum.dim_IntList %1388, %1394, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1396 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1397 = torch.aten.view %1395, %1396 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1398 = torch.aten.detach %1397 : !torch.tensor -> !torch.tensor
    %1399 = torch.aten.detach %1398 : !torch.tensor -> !torch.tensor
    %1400 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1401 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1402 = torch.aten._reshape_alias %1390, %1400, %1401 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1403 = torch.aten.t %1393 : !torch.tensor -> !torch.tensor
    %1404 = torch.aten.detach %1403 : !torch.tensor -> !torch.tensor
    %1405 = torch.aten.detach %1404 : !torch.tensor -> !torch.tensor
    %1406 = torch.aten.gelu_backward %1402, %994, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1407 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %1408 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1409 = torch.aten._reshape_alias %1406, %1407, %1408 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1410 = torch.aten.t %989 : !torch.tensor -> !torch.tensor
    %1411 = torch.aten.mm %1409, %1410 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1412 = torch.aten.t %1409 : !torch.tensor -> !torch.tensor
    %1413 = torch.aten.mm %1412, %991 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1414 = torch.aten.t %1413 : !torch.tensor -> !torch.tensor
    %1415 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1416 = torch.aten.sum.dim_IntList %1409, %1415, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1417 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %1418 = torch.aten.view %1416, %1417 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1419 = torch.aten.detach %1418 : !torch.tensor -> !torch.tensor
    %1420 = torch.aten.detach %1419 : !torch.tensor -> !torch.tensor
    %1421 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1422 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1423 = torch.aten._reshape_alias %1411, %1421, %1422 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1424 = torch.aten.add.Tensor %1376, %1423, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1425 = torch.aten.t %1414 : !torch.tensor -> !torch.tensor
    %1426 = torch.aten.detach %1425 : !torch.tensor -> !torch.tensor
    %1427 = torch.aten.detach %1426 : !torch.tensor -> !torch.tensor
    %1428 = torch.aten.sub.Tensor %987, %result1_64, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1429 = torch.aten.mul.Tensor %1428, %result2_65 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1430 = torch.aten.mul.Tensor %1424, %arg39 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1431 = torch.aten.mul.Scalar %1430, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1432 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1433 = torch.aten.sum.dim_IntList %1430, %1432, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1434 = torch.aten.mul.Tensor %1430, %1429 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1435 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1436 = torch.aten.sum.dim_IntList %1434, %1435, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1437 = torch.aten.mul.Tensor %1429, %1436 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1438 = torch.aten.sub.Tensor %1431, %1433, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1439 = torch.aten.sub.Tensor %1438, %1437, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1440 = torch.aten.div.Scalar %result2_65, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1441 = torch.aten.mul.Tensor %1440, %1439 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1442 = torch.aten.mul.Tensor %1424, %1429 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1443 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1444 = torch.aten.sum.dim_IntList %1442, %1443, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1445 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1446 = torch.aten.sum.dim_IntList %1424, %1445, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1447 = torch.aten.detach %1444 : !torch.tensor -> !torch.tensor
    %1448 = torch.aten.detach %1447 : !torch.tensor -> !torch.tensor
    %1449 = torch.aten.detach %1446 : !torch.tensor -> !torch.tensor
    %1450 = torch.aten.detach %1449 : !torch.tensor -> !torch.tensor
    %1451 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1452 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1453 = torch.aten._reshape_alias %1441, %1451, %1452 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1454 = torch.aten.t %981 : !torch.tensor -> !torch.tensor
    %1455 = torch.aten.mm %1453, %1454 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1456 = torch.aten.t %1453 : !torch.tensor -> !torch.tensor
    %1457 = torch.aten.mm %1456, %983 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1458 = torch.aten.t %1457 : !torch.tensor -> !torch.tensor
    %1459 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1460 = torch.aten.sum.dim_IntList %1453, %1459, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1461 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1462 = torch.aten.view %1460, %1461 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1463 = torch.aten.detach %1462 : !torch.tensor -> !torch.tensor
    %1464 = torch.aten.detach %1463 : !torch.tensor -> !torch.tensor
    %1465 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1466 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1467 = torch.aten._reshape_alias %1455, %1465, %1466 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1468 = torch.aten.t %1458 : !torch.tensor -> !torch.tensor
    %1469 = torch.aten.detach %1468 : !torch.tensor -> !torch.tensor
    %1470 = torch.aten.detach %1469 : !torch.tensor -> !torch.tensor
    %1471 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1472 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1473 = torch.aten._reshape_alias %1467, %1471, %1472 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1474 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1475 = torch.aten.permute %1473, %1474 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1476 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1477 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1478 = torch.aten._reshape_alias %1475, %1476, %1477 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1479 = torch.aten.transpose.int %967, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1480 = torch.aten.bmm %1479, %1478 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1481 = torch.aten.transpose.int %972, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1482 = torch.aten.bmm %1478, %1481 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1483 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1484 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1485 = torch.aten._reshape_alias %1480, %1483, %1484 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1486 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1487 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1488 = torch.aten._reshape_alias %1482, %1486, %1487 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1489 = torch.aten.detach %962 : !torch.tensor -> !torch.tensor
    %1490 = torch.aten._softmax_backward_data %1488, %1489, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1491 = torch.aten.div.Scalar %1490, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %1492 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1493 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1494 = torch.aten._reshape_alias %1491, %1492, %1493 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1495 = torch.aten.transpose.int %949, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1496 = torch.aten.bmm %1495, %1494 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1497 = torch.aten.transpose.int %954, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1498 = torch.aten.bmm %1494, %1497 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1499 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1500 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1501 = torch.aten._reshape_alias %1496, %1499, %1500 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1502 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1503 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1504 = torch.aten._reshape_alias %1498, %1502, %1503 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1505 = torch.aten.transpose.int %1501, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1506 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1507 = torch.aten.permute %1504, %1506 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1508 = torch.aten.clone %1507, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1509 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1510 = torch.aten._unsafe_view %1508, %1509 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1511 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1512 = torch.aten.permute %1485, %1511 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1513 = torch.aten.clone %1512, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1514 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1515 = torch.aten._unsafe_view %1513, %1514 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1516 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1517 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1518 = torch.aten._reshape_alias %1515, %1516, %1517 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1519 = torch.aten.t %930 : !torch.tensor -> !torch.tensor
    %1520 = torch.aten.mm %1518, %1519 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1521 = torch.aten.t %1518 : !torch.tensor -> !torch.tensor
    %1522 = torch.aten.mm %1521, %932 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1523 = torch.aten.t %1522 : !torch.tensor -> !torch.tensor
    %1524 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1525 = torch.aten.sum.dim_IntList %1518, %1524, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1526 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1527 = torch.aten.view %1525, %1526 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1528 = torch.aten.detach %1527 : !torch.tensor -> !torch.tensor
    %1529 = torch.aten.detach %1528 : !torch.tensor -> !torch.tensor
    %1530 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1531 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1532 = torch.aten._reshape_alias %1520, %1530, %1531 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1533 = torch.aten.add.Tensor %1441, %1532, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1534 = torch.aten.t %1523 : !torch.tensor -> !torch.tensor
    %1535 = torch.aten.detach %1534 : !torch.tensor -> !torch.tensor
    %1536 = torch.aten.detach %1535 : !torch.tensor -> !torch.tensor
    %1537 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1538 = torch.aten.permute %1505, %1537 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1539 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1540 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1541 = torch.aten._reshape_alias %1538, %1539, %1540 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1542 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1543 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %1544 = torch.aten._reshape_alias %1541, %1542, %1543 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1545 = torch.aten.t %920 : !torch.tensor -> !torch.tensor
    %1546 = torch.aten.mm %1544, %1545 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1547 = torch.aten.t %1544 : !torch.tensor -> !torch.tensor
    %1548 = torch.aten.mm %1547, %922 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1549 = torch.aten.t %1548 : !torch.tensor -> !torch.tensor
    %1550 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1551 = torch.aten.sum.dim_IntList %1544, %1550, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1552 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1553 = torch.aten.view %1551, %1552 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1554 = torch.aten.detach %1553 : !torch.tensor -> !torch.tensor
    %1555 = torch.aten.detach %1554 : !torch.tensor -> !torch.tensor
    %1556 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1557 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1558 = torch.aten._reshape_alias %1546, %1556, %1557 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1559 = torch.aten.add.Tensor %1533, %1558, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1560 = torch.aten.t %1549 : !torch.tensor -> !torch.tensor
    %1561 = torch.aten.detach %1560 : !torch.tensor -> !torch.tensor
    %1562 = torch.aten.detach %1561 : !torch.tensor -> !torch.tensor
    %1563 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1564 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1565 = torch.aten._reshape_alias %1510, %1563, %1564 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1566 = torch.aten.t %914 : !torch.tensor -> !torch.tensor
    %1567 = torch.aten.mm %1565, %1566 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1568 = torch.aten.t %1565 : !torch.tensor -> !torch.tensor
    %1569 = torch.aten.mm %1568, %916 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1570 = torch.aten.t %1569 : !torch.tensor -> !torch.tensor
    %1571 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1572 = torch.aten.sum.dim_IntList %1565, %1571, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1573 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1574 = torch.aten.view %1572, %1573 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1575 = torch.aten.detach %1574 : !torch.tensor -> !torch.tensor
    %1576 = torch.aten.detach %1575 : !torch.tensor -> !torch.tensor
    %1577 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1578 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1579 = torch.aten._reshape_alias %1567, %1577, %1578 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1580 = torch.aten.add.Tensor %1559, %1579, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1581 = torch.aten.t %1570 : !torch.tensor -> !torch.tensor
    %1582 = torch.aten.detach %1581 : !torch.tensor -> !torch.tensor
    %1583 = torch.aten.detach %1582 : !torch.tensor -> !torch.tensor
    %1584 = torch.aten.sub.Tensor %912, %result1_61, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1585 = torch.aten.mul.Tensor %1584, %result2_62 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1586 = torch.aten.mul.Tensor %1580, %arg195 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1587 = torch.aten.mul.Scalar %1586, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1588 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1589 = torch.aten.sum.dim_IntList %1586, %1588, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1590 = torch.aten.mul.Tensor %1586, %1585 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1591 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1592 = torch.aten.sum.dim_IntList %1590, %1591, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1593 = torch.aten.mul.Tensor %1585, %1592 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1594 = torch.aten.sub.Tensor %1587, %1589, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1595 = torch.aten.sub.Tensor %1594, %1593, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1596 = torch.aten.div.Scalar %result2_62, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1597 = torch.aten.mul.Tensor %1596, %1595 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1598 = torch.aten.mul.Tensor %1580, %1585 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1599 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1600 = torch.aten.sum.dim_IntList %1598, %1599, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1601 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1602 = torch.aten.sum.dim_IntList %1580, %1601, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1603 = torch.aten.detach %1600 : !torch.tensor -> !torch.tensor
    %1604 = torch.aten.detach %1603 : !torch.tensor -> !torch.tensor
    %1605 = torch.aten.detach %1602 : !torch.tensor -> !torch.tensor
    %1606 = torch.aten.detach %1605 : !torch.tensor -> !torch.tensor
    %1607 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1608 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1609 = torch.aten._reshape_alias %1597, %1607, %1608 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1610 = torch.aten.t %906 : !torch.tensor -> !torch.tensor
    %1611 = torch.aten.mm %1609, %1610 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1612 = torch.aten.t %1609 : !torch.tensor -> !torch.tensor
    %1613 = torch.aten.mm %1612, %908 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1614 = torch.aten.t %1613 : !torch.tensor -> !torch.tensor
    %1615 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1616 = torch.aten.sum.dim_IntList %1609, %1615, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1617 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1618 = torch.aten.view %1616, %1617 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1619 = torch.aten.detach %1618 : !torch.tensor -> !torch.tensor
    %1620 = torch.aten.detach %1619 : !torch.tensor -> !torch.tensor
    %1621 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1622 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1623 = torch.aten._reshape_alias %1611, %1621, %1622 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1624 = torch.aten.t %1614 : !torch.tensor -> !torch.tensor
    %1625 = torch.aten.detach %1624 : !torch.tensor -> !torch.tensor
    %1626 = torch.aten.detach %1625 : !torch.tensor -> !torch.tensor
    %1627 = torch.aten.gelu_backward %1623, %904, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1628 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %1629 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1630 = torch.aten._reshape_alias %1627, %1628, %1629 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1631 = torch.aten.t %899 : !torch.tensor -> !torch.tensor
    %1632 = torch.aten.mm %1630, %1631 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1633 = torch.aten.t %1630 : !torch.tensor -> !torch.tensor
    %1634 = torch.aten.mm %1633, %901 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1635 = torch.aten.t %1634 : !torch.tensor -> !torch.tensor
    %1636 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1637 = torch.aten.sum.dim_IntList %1630, %1636, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1638 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %1639 = torch.aten.view %1637, %1638 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1640 = torch.aten.detach %1639 : !torch.tensor -> !torch.tensor
    %1641 = torch.aten.detach %1640 : !torch.tensor -> !torch.tensor
    %1642 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1643 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1644 = torch.aten._reshape_alias %1632, %1642, %1643 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1645 = torch.aten.add.Tensor %1597, %1644, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1646 = torch.aten.t %1635 : !torch.tensor -> !torch.tensor
    %1647 = torch.aten.detach %1646 : !torch.tensor -> !torch.tensor
    %1648 = torch.aten.detach %1647 : !torch.tensor -> !torch.tensor
    %1649 = torch.aten.sub.Tensor %897, %result1_58, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1650 = torch.aten.mul.Tensor %1649, %result2_59 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1651 = torch.aten.mul.Tensor %1645, %arg183 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1652 = torch.aten.mul.Scalar %1651, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1653 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1654 = torch.aten.sum.dim_IntList %1651, %1653, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1655 = torch.aten.mul.Tensor %1651, %1650 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1656 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1657 = torch.aten.sum.dim_IntList %1655, %1656, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1658 = torch.aten.mul.Tensor %1650, %1657 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1659 = torch.aten.sub.Tensor %1652, %1654, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1660 = torch.aten.sub.Tensor %1659, %1658, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1661 = torch.aten.div.Scalar %result2_59, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1662 = torch.aten.mul.Tensor %1661, %1660 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1663 = torch.aten.mul.Tensor %1645, %1650 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1664 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1665 = torch.aten.sum.dim_IntList %1663, %1664, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1666 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1667 = torch.aten.sum.dim_IntList %1645, %1666, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1668 = torch.aten.detach %1665 : !torch.tensor -> !torch.tensor
    %1669 = torch.aten.detach %1668 : !torch.tensor -> !torch.tensor
    %1670 = torch.aten.detach %1667 : !torch.tensor -> !torch.tensor
    %1671 = torch.aten.detach %1670 : !torch.tensor -> !torch.tensor
    %1672 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1673 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1674 = torch.aten._reshape_alias %1662, %1672, %1673 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1675 = torch.aten.t %891 : !torch.tensor -> !torch.tensor
    %1676 = torch.aten.mm %1674, %1675 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1677 = torch.aten.t %1674 : !torch.tensor -> !torch.tensor
    %1678 = torch.aten.mm %1677, %893 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1679 = torch.aten.t %1678 : !torch.tensor -> !torch.tensor
    %1680 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1681 = torch.aten.sum.dim_IntList %1674, %1680, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1682 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1683 = torch.aten.view %1681, %1682 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1684 = torch.aten.detach %1683 : !torch.tensor -> !torch.tensor
    %1685 = torch.aten.detach %1684 : !torch.tensor -> !torch.tensor
    %1686 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1687 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1688 = torch.aten._reshape_alias %1676, %1686, %1687 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1689 = torch.aten.t %1679 : !torch.tensor -> !torch.tensor
    %1690 = torch.aten.detach %1689 : !torch.tensor -> !torch.tensor
    %1691 = torch.aten.detach %1690 : !torch.tensor -> !torch.tensor
    %1692 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1693 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1694 = torch.aten._reshape_alias %1688, %1692, %1693 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1695 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1696 = torch.aten.permute %1694, %1695 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1697 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1698 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1699 = torch.aten._reshape_alias %1696, %1697, %1698 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1700 = torch.aten.transpose.int %877, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1701 = torch.aten.bmm %1700, %1699 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1702 = torch.aten.transpose.int %882, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1703 = torch.aten.bmm %1699, %1702 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1704 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1705 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1706 = torch.aten._reshape_alias %1701, %1704, %1705 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1707 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1708 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1709 = torch.aten._reshape_alias %1703, %1707, %1708 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1710 = torch.aten.detach %872 : !torch.tensor -> !torch.tensor
    %1711 = torch.aten._softmax_backward_data %1709, %1710, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1712 = torch.aten.div.Scalar %1711, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %1713 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1714 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1715 = torch.aten._reshape_alias %1712, %1713, %1714 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1716 = torch.aten.transpose.int %859, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1717 = torch.aten.bmm %1716, %1715 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1718 = torch.aten.transpose.int %864, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1719 = torch.aten.bmm %1715, %1718 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1720 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1721 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1722 = torch.aten._reshape_alias %1717, %1720, %1721 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1723 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1724 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1725 = torch.aten._reshape_alias %1719, %1723, %1724 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1726 = torch.aten.transpose.int %1722, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1727 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1728 = torch.aten.permute %1725, %1727 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1729 = torch.aten.clone %1728, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1730 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1731 = torch.aten._unsafe_view %1729, %1730 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1732 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1733 = torch.aten.permute %1706, %1732 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1734 = torch.aten.clone %1733, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1735 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1736 = torch.aten._unsafe_view %1734, %1735 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1737 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1738 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1739 = torch.aten._reshape_alias %1736, %1737, %1738 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1740 = torch.aten.t %840 : !torch.tensor -> !torch.tensor
    %1741 = torch.aten.mm %1739, %1740 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1742 = torch.aten.t %1739 : !torch.tensor -> !torch.tensor
    %1743 = torch.aten.mm %1742, %842 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1744 = torch.aten.t %1743 : !torch.tensor -> !torch.tensor
    %1745 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1746 = torch.aten.sum.dim_IntList %1739, %1745, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1747 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1748 = torch.aten.view %1746, %1747 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1749 = torch.aten.detach %1748 : !torch.tensor -> !torch.tensor
    %1750 = torch.aten.detach %1749 : !torch.tensor -> !torch.tensor
    %1751 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1752 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1753 = torch.aten._reshape_alias %1741, %1751, %1752 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1754 = torch.aten.add.Tensor %1662, %1753, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1755 = torch.aten.t %1744 : !torch.tensor -> !torch.tensor
    %1756 = torch.aten.detach %1755 : !torch.tensor -> !torch.tensor
    %1757 = torch.aten.detach %1756 : !torch.tensor -> !torch.tensor
    %1758 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1759 = torch.aten.permute %1726, %1758 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1760 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1761 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1762 = torch.aten._reshape_alias %1759, %1760, %1761 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1763 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1764 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %1765 = torch.aten._reshape_alias %1762, %1763, %1764 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1766 = torch.aten.t %830 : !torch.tensor -> !torch.tensor
    %1767 = torch.aten.mm %1765, %1766 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1768 = torch.aten.t %1765 : !torch.tensor -> !torch.tensor
    %1769 = torch.aten.mm %1768, %832 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1770 = torch.aten.t %1769 : !torch.tensor -> !torch.tensor
    %1771 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1772 = torch.aten.sum.dim_IntList %1765, %1771, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1773 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1774 = torch.aten.view %1772, %1773 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1775 = torch.aten.detach %1774 : !torch.tensor -> !torch.tensor
    %1776 = torch.aten.detach %1775 : !torch.tensor -> !torch.tensor
    %1777 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1778 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1779 = torch.aten._reshape_alias %1767, %1777, %1778 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1780 = torch.aten.add.Tensor %1754, %1779, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1781 = torch.aten.t %1770 : !torch.tensor -> !torch.tensor
    %1782 = torch.aten.detach %1781 : !torch.tensor -> !torch.tensor
    %1783 = torch.aten.detach %1782 : !torch.tensor -> !torch.tensor
    %1784 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1785 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1786 = torch.aten._reshape_alias %1731, %1784, %1785 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1787 = torch.aten.t %824 : !torch.tensor -> !torch.tensor
    %1788 = torch.aten.mm %1786, %1787 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1789 = torch.aten.t %1786 : !torch.tensor -> !torch.tensor
    %1790 = torch.aten.mm %1789, %826 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1791 = torch.aten.t %1790 : !torch.tensor -> !torch.tensor
    %1792 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1793 = torch.aten.sum.dim_IntList %1786, %1792, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1794 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1795 = torch.aten.view %1793, %1794 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1796 = torch.aten.detach %1795 : !torch.tensor -> !torch.tensor
    %1797 = torch.aten.detach %1796 : !torch.tensor -> !torch.tensor
    %1798 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1799 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1800 = torch.aten._reshape_alias %1788, %1798, %1799 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1801 = torch.aten.add.Tensor %1780, %1800, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1802 = torch.aten.t %1791 : !torch.tensor -> !torch.tensor
    %1803 = torch.aten.detach %1802 : !torch.tensor -> !torch.tensor
    %1804 = torch.aten.detach %1803 : !torch.tensor -> !torch.tensor
    %1805 = torch.aten.sub.Tensor %822, %result1_55, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1806 = torch.aten.mul.Tensor %1805, %result2_56 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1807 = torch.aten.mul.Tensor %1801, %arg179 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1808 = torch.aten.mul.Scalar %1807, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1809 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1810 = torch.aten.sum.dim_IntList %1807, %1809, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1811 = torch.aten.mul.Tensor %1807, %1806 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1812 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1813 = torch.aten.sum.dim_IntList %1811, %1812, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1814 = torch.aten.mul.Tensor %1806, %1813 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1815 = torch.aten.sub.Tensor %1808, %1810, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1816 = torch.aten.sub.Tensor %1815, %1814, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1817 = torch.aten.div.Scalar %result2_56, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1818 = torch.aten.mul.Tensor %1817, %1816 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1819 = torch.aten.mul.Tensor %1801, %1806 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1820 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1821 = torch.aten.sum.dim_IntList %1819, %1820, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1822 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1823 = torch.aten.sum.dim_IntList %1801, %1822, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1824 = torch.aten.detach %1821 : !torch.tensor -> !torch.tensor
    %1825 = torch.aten.detach %1824 : !torch.tensor -> !torch.tensor
    %1826 = torch.aten.detach %1823 : !torch.tensor -> !torch.tensor
    %1827 = torch.aten.detach %1826 : !torch.tensor -> !torch.tensor
    %1828 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1829 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1830 = torch.aten._reshape_alias %1818, %1828, %1829 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1831 = torch.aten.t %816 : !torch.tensor -> !torch.tensor
    %1832 = torch.aten.mm %1830, %1831 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1833 = torch.aten.t %1830 : !torch.tensor -> !torch.tensor
    %1834 = torch.aten.mm %1833, %818 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1835 = torch.aten.t %1834 : !torch.tensor -> !torch.tensor
    %1836 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1837 = torch.aten.sum.dim_IntList %1830, %1836, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1838 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1839 = torch.aten.view %1837, %1838 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1840 = torch.aten.detach %1839 : !torch.tensor -> !torch.tensor
    %1841 = torch.aten.detach %1840 : !torch.tensor -> !torch.tensor
    %1842 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1843 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1844 = torch.aten._reshape_alias %1832, %1842, %1843 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1845 = torch.aten.t %1835 : !torch.tensor -> !torch.tensor
    %1846 = torch.aten.detach %1845 : !torch.tensor -> !torch.tensor
    %1847 = torch.aten.detach %1846 : !torch.tensor -> !torch.tensor
    %1848 = torch.aten.gelu_backward %1844, %814, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1849 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %1850 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1851 = torch.aten._reshape_alias %1848, %1849, %1850 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1852 = torch.aten.t %809 : !torch.tensor -> !torch.tensor
    %1853 = torch.aten.mm %1851, %1852 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1854 = torch.aten.t %1851 : !torch.tensor -> !torch.tensor
    %1855 = torch.aten.mm %1854, %811 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1856 = torch.aten.t %1855 : !torch.tensor -> !torch.tensor
    %1857 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1858 = torch.aten.sum.dim_IntList %1851, %1857, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1859 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %1860 = torch.aten.view %1858, %1859 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1861 = torch.aten.detach %1860 : !torch.tensor -> !torch.tensor
    %1862 = torch.aten.detach %1861 : !torch.tensor -> !torch.tensor
    %1863 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1864 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1865 = torch.aten._reshape_alias %1853, %1863, %1864 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1866 = torch.aten.add.Tensor %1818, %1865, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1867 = torch.aten.t %1856 : !torch.tensor -> !torch.tensor
    %1868 = torch.aten.detach %1867 : !torch.tensor -> !torch.tensor
    %1869 = torch.aten.detach %1868 : !torch.tensor -> !torch.tensor
    %1870 = torch.aten.sub.Tensor %807, %result1_52, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1871 = torch.aten.mul.Tensor %1870, %result2_53 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1872 = torch.aten.mul.Tensor %1866, %arg167 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1873 = torch.aten.mul.Scalar %1872, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1874 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1875 = torch.aten.sum.dim_IntList %1872, %1874, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1876 = torch.aten.mul.Tensor %1872, %1871 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1877 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1878 = torch.aten.sum.dim_IntList %1876, %1877, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1879 = torch.aten.mul.Tensor %1871, %1878 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1880 = torch.aten.sub.Tensor %1873, %1875, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1881 = torch.aten.sub.Tensor %1880, %1879, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1882 = torch.aten.div.Scalar %result2_53, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %1883 = torch.aten.mul.Tensor %1882, %1881 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1884 = torch.aten.mul.Tensor %1866, %1871 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1885 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1886 = torch.aten.sum.dim_IntList %1884, %1885, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1887 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1888 = torch.aten.sum.dim_IntList %1866, %1887, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1889 = torch.aten.detach %1886 : !torch.tensor -> !torch.tensor
    %1890 = torch.aten.detach %1889 : !torch.tensor -> !torch.tensor
    %1891 = torch.aten.detach %1888 : !torch.tensor -> !torch.tensor
    %1892 = torch.aten.detach %1891 : !torch.tensor -> !torch.tensor
    %1893 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1894 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1895 = torch.aten._reshape_alias %1883, %1893, %1894 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1896 = torch.aten.t %801 : !torch.tensor -> !torch.tensor
    %1897 = torch.aten.mm %1895, %1896 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1898 = torch.aten.t %1895 : !torch.tensor -> !torch.tensor
    %1899 = torch.aten.mm %1898, %803 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1900 = torch.aten.t %1899 : !torch.tensor -> !torch.tensor
    %1901 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1902 = torch.aten.sum.dim_IntList %1895, %1901, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1903 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1904 = torch.aten.view %1902, %1903 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1905 = torch.aten.detach %1904 : !torch.tensor -> !torch.tensor
    %1906 = torch.aten.detach %1905 : !torch.tensor -> !torch.tensor
    %1907 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1908 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1909 = torch.aten._reshape_alias %1897, %1907, %1908 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1910 = torch.aten.t %1900 : !torch.tensor -> !torch.tensor
    %1911 = torch.aten.detach %1910 : !torch.tensor -> !torch.tensor
    %1912 = torch.aten.detach %1911 : !torch.tensor -> !torch.tensor
    %1913 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1914 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1915 = torch.aten._reshape_alias %1909, %1913, %1914 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1916 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1917 = torch.aten.permute %1915, %1916 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1918 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1919 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1920 = torch.aten._reshape_alias %1917, %1918, %1919 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1921 = torch.aten.transpose.int %787, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1922 = torch.aten.bmm %1921, %1920 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1923 = torch.aten.transpose.int %792, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1924 = torch.aten.bmm %1920, %1923 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1925 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1926 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1927 = torch.aten._reshape_alias %1922, %1925, %1926 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1928 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1929 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1930 = torch.aten._reshape_alias %1924, %1928, %1929 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1931 = torch.aten.detach %782 : !torch.tensor -> !torch.tensor
    %1932 = torch.aten._softmax_backward_data %1930, %1931, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1933 = torch.aten.div.Scalar %1932, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %1934 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1935 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1936 = torch.aten._reshape_alias %1933, %1934, %1935 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1937 = torch.aten.transpose.int %769, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1938 = torch.aten.bmm %1937, %1936 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1939 = torch.aten.transpose.int %774, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1940 = torch.aten.bmm %1936, %1939 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1941 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1942 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1943 = torch.aten._reshape_alias %1938, %1941, %1942 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1944 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1945 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1946 = torch.aten._reshape_alias %1940, %1944, %1945 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1947 = torch.aten.transpose.int %1943, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1948 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1949 = torch.aten.permute %1946, %1948 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1950 = torch.aten.clone %1949, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1951 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1952 = torch.aten._unsafe_view %1950, %1951 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1953 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1954 = torch.aten.permute %1927, %1953 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1955 = torch.aten.clone %1954, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1956 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1957 = torch.aten._unsafe_view %1955, %1956 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1958 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1959 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1960 = torch.aten._reshape_alias %1957, %1958, %1959 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1961 = torch.aten.t %750 : !torch.tensor -> !torch.tensor
    %1962 = torch.aten.mm %1960, %1961 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1963 = torch.aten.t %1960 : !torch.tensor -> !torch.tensor
    %1964 = torch.aten.mm %1963, %752 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1965 = torch.aten.t %1964 : !torch.tensor -> !torch.tensor
    %1966 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1967 = torch.aten.sum.dim_IntList %1960, %1966, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1968 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1969 = torch.aten.view %1967, %1968 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1970 = torch.aten.detach %1969 : !torch.tensor -> !torch.tensor
    %1971 = torch.aten.detach %1970 : !torch.tensor -> !torch.tensor
    %1972 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1973 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1974 = torch.aten._reshape_alias %1962, %1972, %1973 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1975 = torch.aten.add.Tensor %1883, %1974, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1976 = torch.aten.t %1965 : !torch.tensor -> !torch.tensor
    %1977 = torch.aten.detach %1976 : !torch.tensor -> !torch.tensor
    %1978 = torch.aten.detach %1977 : !torch.tensor -> !torch.tensor
    %1979 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1980 = torch.aten.permute %1947, %1979 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1981 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1982 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1983 = torch.aten._reshape_alias %1980, %1981, %1982 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1984 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %1985 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %1986 = torch.aten._reshape_alias %1983, %1984, %1985 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %1987 = torch.aten.t %740 : !torch.tensor -> !torch.tensor
    %1988 = torch.aten.mm %1986, %1987 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1989 = torch.aten.t %1986 : !torch.tensor -> !torch.tensor
    %1990 = torch.aten.mm %1989, %742 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1991 = torch.aten.t %1990 : !torch.tensor -> !torch.tensor
    %1992 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1993 = torch.aten.sum.dim_IntList %1986, %1992, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1994 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %1995 = torch.aten.view %1993, %1994 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1996 = torch.aten.detach %1995 : !torch.tensor -> !torch.tensor
    %1997 = torch.aten.detach %1996 : !torch.tensor -> !torch.tensor
    %1998 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1999 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2000 = torch.aten._reshape_alias %1988, %1998, %1999 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2001 = torch.aten.add.Tensor %1975, %2000, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2002 = torch.aten.t %1991 : !torch.tensor -> !torch.tensor
    %2003 = torch.aten.detach %2002 : !torch.tensor -> !torch.tensor
    %2004 = torch.aten.detach %2003 : !torch.tensor -> !torch.tensor
    %2005 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2006 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2007 = torch.aten._reshape_alias %1952, %2005, %2006 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2008 = torch.aten.t %734 : !torch.tensor -> !torch.tensor
    %2009 = torch.aten.mm %2007, %2008 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2010 = torch.aten.t %2007 : !torch.tensor -> !torch.tensor
    %2011 = torch.aten.mm %2010, %736 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2012 = torch.aten.t %2011 : !torch.tensor -> !torch.tensor
    %2013 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2014 = torch.aten.sum.dim_IntList %2007, %2013, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2015 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2016 = torch.aten.view %2014, %2015 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2017 = torch.aten.detach %2016 : !torch.tensor -> !torch.tensor
    %2018 = torch.aten.detach %2017 : !torch.tensor -> !torch.tensor
    %2019 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2020 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2021 = torch.aten._reshape_alias %2009, %2019, %2020 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2022 = torch.aten.add.Tensor %2001, %2021, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2023 = torch.aten.t %2012 : !torch.tensor -> !torch.tensor
    %2024 = torch.aten.detach %2023 : !torch.tensor -> !torch.tensor
    %2025 = torch.aten.detach %2024 : !torch.tensor -> !torch.tensor
    %2026 = torch.aten.sub.Tensor %732, %result1_49, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2027 = torch.aten.mul.Tensor %2026, %result2_50 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2028 = torch.aten.mul.Tensor %2022, %arg163 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2029 = torch.aten.mul.Scalar %2028, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2030 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2031 = torch.aten.sum.dim_IntList %2028, %2030, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2032 = torch.aten.mul.Tensor %2028, %2027 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2033 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2034 = torch.aten.sum.dim_IntList %2032, %2033, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2035 = torch.aten.mul.Tensor %2027, %2034 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2036 = torch.aten.sub.Tensor %2029, %2031, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2037 = torch.aten.sub.Tensor %2036, %2035, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2038 = torch.aten.div.Scalar %result2_50, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2039 = torch.aten.mul.Tensor %2038, %2037 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2040 = torch.aten.mul.Tensor %2022, %2027 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2041 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2042 = torch.aten.sum.dim_IntList %2040, %2041, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2043 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2044 = torch.aten.sum.dim_IntList %2022, %2043, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2045 = torch.aten.detach %2042 : !torch.tensor -> !torch.tensor
    %2046 = torch.aten.detach %2045 : !torch.tensor -> !torch.tensor
    %2047 = torch.aten.detach %2044 : !torch.tensor -> !torch.tensor
    %2048 = torch.aten.detach %2047 : !torch.tensor -> !torch.tensor
    %2049 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2050 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2051 = torch.aten._reshape_alias %2039, %2049, %2050 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2052 = torch.aten.t %726 : !torch.tensor -> !torch.tensor
    %2053 = torch.aten.mm %2051, %2052 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2054 = torch.aten.t %2051 : !torch.tensor -> !torch.tensor
    %2055 = torch.aten.mm %2054, %728 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2056 = torch.aten.t %2055 : !torch.tensor -> !torch.tensor
    %2057 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2058 = torch.aten.sum.dim_IntList %2051, %2057, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2059 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2060 = torch.aten.view %2058, %2059 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2061 = torch.aten.detach %2060 : !torch.tensor -> !torch.tensor
    %2062 = torch.aten.detach %2061 : !torch.tensor -> !torch.tensor
    %2063 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2064 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2065 = torch.aten._reshape_alias %2053, %2063, %2064 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2066 = torch.aten.t %2056 : !torch.tensor -> !torch.tensor
    %2067 = torch.aten.detach %2066 : !torch.tensor -> !torch.tensor
    %2068 = torch.aten.detach %2067 : !torch.tensor -> !torch.tensor
    %2069 = torch.aten.gelu_backward %2065, %724, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2070 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %2071 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2072 = torch.aten._reshape_alias %2069, %2070, %2071 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2073 = torch.aten.t %719 : !torch.tensor -> !torch.tensor
    %2074 = torch.aten.mm %2072, %2073 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2075 = torch.aten.t %2072 : !torch.tensor -> !torch.tensor
    %2076 = torch.aten.mm %2075, %721 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2077 = torch.aten.t %2076 : !torch.tensor -> !torch.tensor
    %2078 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2079 = torch.aten.sum.dim_IntList %2072, %2078, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2080 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %2081 = torch.aten.view %2079, %2080 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2082 = torch.aten.detach %2081 : !torch.tensor -> !torch.tensor
    %2083 = torch.aten.detach %2082 : !torch.tensor -> !torch.tensor
    %2084 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2085 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2086 = torch.aten._reshape_alias %2074, %2084, %2085 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2087 = torch.aten.add.Tensor %2039, %2086, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2088 = torch.aten.t %2077 : !torch.tensor -> !torch.tensor
    %2089 = torch.aten.detach %2088 : !torch.tensor -> !torch.tensor
    %2090 = torch.aten.detach %2089 : !torch.tensor -> !torch.tensor
    %2091 = torch.aten.sub.Tensor %717, %result1_46, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2092 = torch.aten.mul.Tensor %2091, %result2_47 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2093 = torch.aten.mul.Tensor %2087, %arg151 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2094 = torch.aten.mul.Scalar %2093, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2095 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2096 = torch.aten.sum.dim_IntList %2093, %2095, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2097 = torch.aten.mul.Tensor %2093, %2092 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2098 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2099 = torch.aten.sum.dim_IntList %2097, %2098, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2100 = torch.aten.mul.Tensor %2092, %2099 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2101 = torch.aten.sub.Tensor %2094, %2096, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2102 = torch.aten.sub.Tensor %2101, %2100, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2103 = torch.aten.div.Scalar %result2_47, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2104 = torch.aten.mul.Tensor %2103, %2102 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2105 = torch.aten.mul.Tensor %2087, %2092 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2106 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2107 = torch.aten.sum.dim_IntList %2105, %2106, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2108 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2109 = torch.aten.sum.dim_IntList %2087, %2108, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2110 = torch.aten.detach %2107 : !torch.tensor -> !torch.tensor
    %2111 = torch.aten.detach %2110 : !torch.tensor -> !torch.tensor
    %2112 = torch.aten.detach %2109 : !torch.tensor -> !torch.tensor
    %2113 = torch.aten.detach %2112 : !torch.tensor -> !torch.tensor
    %2114 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2115 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2116 = torch.aten._reshape_alias %2104, %2114, %2115 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2117 = torch.aten.t %711 : !torch.tensor -> !torch.tensor
    %2118 = torch.aten.mm %2116, %2117 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2119 = torch.aten.t %2116 : !torch.tensor -> !torch.tensor
    %2120 = torch.aten.mm %2119, %713 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2121 = torch.aten.t %2120 : !torch.tensor -> !torch.tensor
    %2122 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2123 = torch.aten.sum.dim_IntList %2116, %2122, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2124 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2125 = torch.aten.view %2123, %2124 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2126 = torch.aten.detach %2125 : !torch.tensor -> !torch.tensor
    %2127 = torch.aten.detach %2126 : !torch.tensor -> !torch.tensor
    %2128 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2129 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2130 = torch.aten._reshape_alias %2118, %2128, %2129 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2131 = torch.aten.t %2121 : !torch.tensor -> !torch.tensor
    %2132 = torch.aten.detach %2131 : !torch.tensor -> !torch.tensor
    %2133 = torch.aten.detach %2132 : !torch.tensor -> !torch.tensor
    %2134 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2135 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2136 = torch.aten._reshape_alias %2130, %2134, %2135 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2137 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2138 = torch.aten.permute %2136, %2137 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2139 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2140 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2141 = torch.aten._reshape_alias %2138, %2139, %2140 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2142 = torch.aten.transpose.int %697, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2143 = torch.aten.bmm %2142, %2141 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2144 = torch.aten.transpose.int %702, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2145 = torch.aten.bmm %2141, %2144 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2146 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2147 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2148 = torch.aten._reshape_alias %2143, %2146, %2147 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2149 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2150 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2151 = torch.aten._reshape_alias %2145, %2149, %2150 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2152 = torch.aten.detach %692 : !torch.tensor -> !torch.tensor
    %2153 = torch.aten._softmax_backward_data %2151, %2152, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2154 = torch.aten.div.Scalar %2153, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %2155 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2156 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2157 = torch.aten._reshape_alias %2154, %2155, %2156 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2158 = torch.aten.transpose.int %679, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2159 = torch.aten.bmm %2158, %2157 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2160 = torch.aten.transpose.int %684, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2161 = torch.aten.bmm %2157, %2160 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2162 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2163 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2164 = torch.aten._reshape_alias %2159, %2162, %2163 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2165 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2166 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2167 = torch.aten._reshape_alias %2161, %2165, %2166 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2168 = torch.aten.transpose.int %2164, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2169 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2170 = torch.aten.permute %2167, %2169 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2171 = torch.aten.clone %2170, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2172 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2173 = torch.aten._unsafe_view %2171, %2172 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2174 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2175 = torch.aten.permute %2148, %2174 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2176 = torch.aten.clone %2175, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2177 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2178 = torch.aten._unsafe_view %2176, %2177 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2179 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2180 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2181 = torch.aten._reshape_alias %2178, %2179, %2180 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2182 = torch.aten.t %660 : !torch.tensor -> !torch.tensor
    %2183 = torch.aten.mm %2181, %2182 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2184 = torch.aten.t %2181 : !torch.tensor -> !torch.tensor
    %2185 = torch.aten.mm %2184, %662 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2186 = torch.aten.t %2185 : !torch.tensor -> !torch.tensor
    %2187 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2188 = torch.aten.sum.dim_IntList %2181, %2187, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2189 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2190 = torch.aten.view %2188, %2189 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2191 = torch.aten.detach %2190 : !torch.tensor -> !torch.tensor
    %2192 = torch.aten.detach %2191 : !torch.tensor -> !torch.tensor
    %2193 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2194 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2195 = torch.aten._reshape_alias %2183, %2193, %2194 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2196 = torch.aten.add.Tensor %2104, %2195, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2197 = torch.aten.t %2186 : !torch.tensor -> !torch.tensor
    %2198 = torch.aten.detach %2197 : !torch.tensor -> !torch.tensor
    %2199 = torch.aten.detach %2198 : !torch.tensor -> !torch.tensor
    %2200 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2201 = torch.aten.permute %2168, %2200 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2202 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2203 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2204 = torch.aten._reshape_alias %2201, %2202, %2203 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2205 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2206 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %2207 = torch.aten._reshape_alias %2204, %2205, %2206 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2208 = torch.aten.t %650 : !torch.tensor -> !torch.tensor
    %2209 = torch.aten.mm %2207, %2208 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2210 = torch.aten.t %2207 : !torch.tensor -> !torch.tensor
    %2211 = torch.aten.mm %2210, %652 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2212 = torch.aten.t %2211 : !torch.tensor -> !torch.tensor
    %2213 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2214 = torch.aten.sum.dim_IntList %2207, %2213, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2215 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2216 = torch.aten.view %2214, %2215 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2217 = torch.aten.detach %2216 : !torch.tensor -> !torch.tensor
    %2218 = torch.aten.detach %2217 : !torch.tensor -> !torch.tensor
    %2219 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2220 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2221 = torch.aten._reshape_alias %2209, %2219, %2220 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2222 = torch.aten.add.Tensor %2196, %2221, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2223 = torch.aten.t %2212 : !torch.tensor -> !torch.tensor
    %2224 = torch.aten.detach %2223 : !torch.tensor -> !torch.tensor
    %2225 = torch.aten.detach %2224 : !torch.tensor -> !torch.tensor
    %2226 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2227 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2228 = torch.aten._reshape_alias %2173, %2226, %2227 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2229 = torch.aten.t %644 : !torch.tensor -> !torch.tensor
    %2230 = torch.aten.mm %2228, %2229 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2231 = torch.aten.t %2228 : !torch.tensor -> !torch.tensor
    %2232 = torch.aten.mm %2231, %646 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2233 = torch.aten.t %2232 : !torch.tensor -> !torch.tensor
    %2234 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2235 = torch.aten.sum.dim_IntList %2228, %2234, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2236 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2237 = torch.aten.view %2235, %2236 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2238 = torch.aten.detach %2237 : !torch.tensor -> !torch.tensor
    %2239 = torch.aten.detach %2238 : !torch.tensor -> !torch.tensor
    %2240 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2241 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2242 = torch.aten._reshape_alias %2230, %2240, %2241 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2243 = torch.aten.add.Tensor %2222, %2242, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2244 = torch.aten.t %2233 : !torch.tensor -> !torch.tensor
    %2245 = torch.aten.detach %2244 : !torch.tensor -> !torch.tensor
    %2246 = torch.aten.detach %2245 : !torch.tensor -> !torch.tensor
    %2247 = torch.aten.sub.Tensor %642, %result1_43, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2248 = torch.aten.mul.Tensor %2247, %result2_44 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2249 = torch.aten.mul.Tensor %2243, %arg147 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2250 = torch.aten.mul.Scalar %2249, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2251 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2252 = torch.aten.sum.dim_IntList %2249, %2251, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2253 = torch.aten.mul.Tensor %2249, %2248 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2254 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2255 = torch.aten.sum.dim_IntList %2253, %2254, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2256 = torch.aten.mul.Tensor %2248, %2255 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2257 = torch.aten.sub.Tensor %2250, %2252, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2258 = torch.aten.sub.Tensor %2257, %2256, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2259 = torch.aten.div.Scalar %result2_44, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2260 = torch.aten.mul.Tensor %2259, %2258 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2261 = torch.aten.mul.Tensor %2243, %2248 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2262 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2263 = torch.aten.sum.dim_IntList %2261, %2262, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2264 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2265 = torch.aten.sum.dim_IntList %2243, %2264, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2266 = torch.aten.detach %2263 : !torch.tensor -> !torch.tensor
    %2267 = torch.aten.detach %2266 : !torch.tensor -> !torch.tensor
    %2268 = torch.aten.detach %2265 : !torch.tensor -> !torch.tensor
    %2269 = torch.aten.detach %2268 : !torch.tensor -> !torch.tensor
    %2270 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2271 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2272 = torch.aten._reshape_alias %2260, %2270, %2271 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2273 = torch.aten.t %636 : !torch.tensor -> !torch.tensor
    %2274 = torch.aten.mm %2272, %2273 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2275 = torch.aten.t %2272 : !torch.tensor -> !torch.tensor
    %2276 = torch.aten.mm %2275, %638 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2277 = torch.aten.t %2276 : !torch.tensor -> !torch.tensor
    %2278 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2279 = torch.aten.sum.dim_IntList %2272, %2278, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2280 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2281 = torch.aten.view %2279, %2280 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2282 = torch.aten.detach %2281 : !torch.tensor -> !torch.tensor
    %2283 = torch.aten.detach %2282 : !torch.tensor -> !torch.tensor
    %2284 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2285 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2286 = torch.aten._reshape_alias %2274, %2284, %2285 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2287 = torch.aten.t %2277 : !torch.tensor -> !torch.tensor
    %2288 = torch.aten.detach %2287 : !torch.tensor -> !torch.tensor
    %2289 = torch.aten.detach %2288 : !torch.tensor -> !torch.tensor
    %2290 = torch.aten.gelu_backward %2286, %634, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2291 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %2292 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2293 = torch.aten._reshape_alias %2290, %2291, %2292 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2294 = torch.aten.t %629 : !torch.tensor -> !torch.tensor
    %2295 = torch.aten.mm %2293, %2294 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2296 = torch.aten.t %2293 : !torch.tensor -> !torch.tensor
    %2297 = torch.aten.mm %2296, %631 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2298 = torch.aten.t %2297 : !torch.tensor -> !torch.tensor
    %2299 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2300 = torch.aten.sum.dim_IntList %2293, %2299, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2301 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %2302 = torch.aten.view %2300, %2301 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2303 = torch.aten.detach %2302 : !torch.tensor -> !torch.tensor
    %2304 = torch.aten.detach %2303 : !torch.tensor -> !torch.tensor
    %2305 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2306 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2307 = torch.aten._reshape_alias %2295, %2305, %2306 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2308 = torch.aten.add.Tensor %2260, %2307, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2309 = torch.aten.t %2298 : !torch.tensor -> !torch.tensor
    %2310 = torch.aten.detach %2309 : !torch.tensor -> !torch.tensor
    %2311 = torch.aten.detach %2310 : !torch.tensor -> !torch.tensor
    %2312 = torch.aten.sub.Tensor %627, %result1_40, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2313 = torch.aten.mul.Tensor %2312, %result2_41 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2314 = torch.aten.mul.Tensor %2308, %arg135 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2315 = torch.aten.mul.Scalar %2314, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2316 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2317 = torch.aten.sum.dim_IntList %2314, %2316, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2318 = torch.aten.mul.Tensor %2314, %2313 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2319 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2320 = torch.aten.sum.dim_IntList %2318, %2319, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2321 = torch.aten.mul.Tensor %2313, %2320 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2322 = torch.aten.sub.Tensor %2315, %2317, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2323 = torch.aten.sub.Tensor %2322, %2321, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2324 = torch.aten.div.Scalar %result2_41, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2325 = torch.aten.mul.Tensor %2324, %2323 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2326 = torch.aten.mul.Tensor %2308, %2313 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2327 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2328 = torch.aten.sum.dim_IntList %2326, %2327, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2329 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2330 = torch.aten.sum.dim_IntList %2308, %2329, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2331 = torch.aten.detach %2328 : !torch.tensor -> !torch.tensor
    %2332 = torch.aten.detach %2331 : !torch.tensor -> !torch.tensor
    %2333 = torch.aten.detach %2330 : !torch.tensor -> !torch.tensor
    %2334 = torch.aten.detach %2333 : !torch.tensor -> !torch.tensor
    %2335 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2336 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2337 = torch.aten._reshape_alias %2325, %2335, %2336 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2338 = torch.aten.t %621 : !torch.tensor -> !torch.tensor
    %2339 = torch.aten.mm %2337, %2338 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2340 = torch.aten.t %2337 : !torch.tensor -> !torch.tensor
    %2341 = torch.aten.mm %2340, %623 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2342 = torch.aten.t %2341 : !torch.tensor -> !torch.tensor
    %2343 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2344 = torch.aten.sum.dim_IntList %2337, %2343, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2345 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2346 = torch.aten.view %2344, %2345 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2347 = torch.aten.detach %2346 : !torch.tensor -> !torch.tensor
    %2348 = torch.aten.detach %2347 : !torch.tensor -> !torch.tensor
    %2349 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2350 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2351 = torch.aten._reshape_alias %2339, %2349, %2350 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2352 = torch.aten.t %2342 : !torch.tensor -> !torch.tensor
    %2353 = torch.aten.detach %2352 : !torch.tensor -> !torch.tensor
    %2354 = torch.aten.detach %2353 : !torch.tensor -> !torch.tensor
    %2355 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2356 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2357 = torch.aten._reshape_alias %2351, %2355, %2356 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2358 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2359 = torch.aten.permute %2357, %2358 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2360 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2361 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2362 = torch.aten._reshape_alias %2359, %2360, %2361 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2363 = torch.aten.transpose.int %607, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2364 = torch.aten.bmm %2363, %2362 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2365 = torch.aten.transpose.int %612, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2366 = torch.aten.bmm %2362, %2365 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2367 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2368 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2369 = torch.aten._reshape_alias %2364, %2367, %2368 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2370 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2371 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2372 = torch.aten._reshape_alias %2366, %2370, %2371 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2373 = torch.aten.detach %602 : !torch.tensor -> !torch.tensor
    %2374 = torch.aten._softmax_backward_data %2372, %2373, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2375 = torch.aten.div.Scalar %2374, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %2376 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2377 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2378 = torch.aten._reshape_alias %2375, %2376, %2377 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2379 = torch.aten.transpose.int %589, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2380 = torch.aten.bmm %2379, %2378 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2381 = torch.aten.transpose.int %594, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2382 = torch.aten.bmm %2378, %2381 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2383 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2384 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2385 = torch.aten._reshape_alias %2380, %2383, %2384 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2386 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2387 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2388 = torch.aten._reshape_alias %2382, %2386, %2387 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2389 = torch.aten.transpose.int %2385, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2390 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2391 = torch.aten.permute %2388, %2390 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2392 = torch.aten.clone %2391, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2393 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2394 = torch.aten._unsafe_view %2392, %2393 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2395 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2396 = torch.aten.permute %2369, %2395 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2397 = torch.aten.clone %2396, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2398 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2399 = torch.aten._unsafe_view %2397, %2398 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2400 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2401 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2402 = torch.aten._reshape_alias %2399, %2400, %2401 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2403 = torch.aten.t %570 : !torch.tensor -> !torch.tensor
    %2404 = torch.aten.mm %2402, %2403 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2405 = torch.aten.t %2402 : !torch.tensor -> !torch.tensor
    %2406 = torch.aten.mm %2405, %572 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2407 = torch.aten.t %2406 : !torch.tensor -> !torch.tensor
    %2408 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2409 = torch.aten.sum.dim_IntList %2402, %2408, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2410 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2411 = torch.aten.view %2409, %2410 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2412 = torch.aten.detach %2411 : !torch.tensor -> !torch.tensor
    %2413 = torch.aten.detach %2412 : !torch.tensor -> !torch.tensor
    %2414 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2415 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2416 = torch.aten._reshape_alias %2404, %2414, %2415 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2417 = torch.aten.add.Tensor %2325, %2416, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2418 = torch.aten.t %2407 : !torch.tensor -> !torch.tensor
    %2419 = torch.aten.detach %2418 : !torch.tensor -> !torch.tensor
    %2420 = torch.aten.detach %2419 : !torch.tensor -> !torch.tensor
    %2421 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2422 = torch.aten.permute %2389, %2421 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2423 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2424 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2425 = torch.aten._reshape_alias %2422, %2423, %2424 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2426 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2427 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %2428 = torch.aten._reshape_alias %2425, %2426, %2427 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2429 = torch.aten.t %560 : !torch.tensor -> !torch.tensor
    %2430 = torch.aten.mm %2428, %2429 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2431 = torch.aten.t %2428 : !torch.tensor -> !torch.tensor
    %2432 = torch.aten.mm %2431, %562 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2433 = torch.aten.t %2432 : !torch.tensor -> !torch.tensor
    %2434 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2435 = torch.aten.sum.dim_IntList %2428, %2434, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2436 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2437 = torch.aten.view %2435, %2436 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2438 = torch.aten.detach %2437 : !torch.tensor -> !torch.tensor
    %2439 = torch.aten.detach %2438 : !torch.tensor -> !torch.tensor
    %2440 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2441 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2442 = torch.aten._reshape_alias %2430, %2440, %2441 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2443 = torch.aten.add.Tensor %2417, %2442, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2444 = torch.aten.t %2433 : !torch.tensor -> !torch.tensor
    %2445 = torch.aten.detach %2444 : !torch.tensor -> !torch.tensor
    %2446 = torch.aten.detach %2445 : !torch.tensor -> !torch.tensor
    %2447 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2448 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2449 = torch.aten._reshape_alias %2394, %2447, %2448 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2450 = torch.aten.t %554 : !torch.tensor -> !torch.tensor
    %2451 = torch.aten.mm %2449, %2450 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2452 = torch.aten.t %2449 : !torch.tensor -> !torch.tensor
    %2453 = torch.aten.mm %2452, %556 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2454 = torch.aten.t %2453 : !torch.tensor -> !torch.tensor
    %2455 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2456 = torch.aten.sum.dim_IntList %2449, %2455, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2457 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2458 = torch.aten.view %2456, %2457 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2459 = torch.aten.detach %2458 : !torch.tensor -> !torch.tensor
    %2460 = torch.aten.detach %2459 : !torch.tensor -> !torch.tensor
    %2461 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2462 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2463 = torch.aten._reshape_alias %2451, %2461, %2462 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2464 = torch.aten.add.Tensor %2443, %2463, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2465 = torch.aten.t %2454 : !torch.tensor -> !torch.tensor
    %2466 = torch.aten.detach %2465 : !torch.tensor -> !torch.tensor
    %2467 = torch.aten.detach %2466 : !torch.tensor -> !torch.tensor
    %2468 = torch.aten.sub.Tensor %552, %result1_37, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2469 = torch.aten.mul.Tensor %2468, %result2_38 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2470 = torch.aten.mul.Tensor %2464, %arg131 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2471 = torch.aten.mul.Scalar %2470, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2472 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2473 = torch.aten.sum.dim_IntList %2470, %2472, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2474 = torch.aten.mul.Tensor %2470, %2469 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2475 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2476 = torch.aten.sum.dim_IntList %2474, %2475, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2477 = torch.aten.mul.Tensor %2469, %2476 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2478 = torch.aten.sub.Tensor %2471, %2473, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2479 = torch.aten.sub.Tensor %2478, %2477, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2480 = torch.aten.div.Scalar %result2_38, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2481 = torch.aten.mul.Tensor %2480, %2479 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2482 = torch.aten.mul.Tensor %2464, %2469 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2483 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2484 = torch.aten.sum.dim_IntList %2482, %2483, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2485 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2486 = torch.aten.sum.dim_IntList %2464, %2485, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2487 = torch.aten.detach %2484 : !torch.tensor -> !torch.tensor
    %2488 = torch.aten.detach %2487 : !torch.tensor -> !torch.tensor
    %2489 = torch.aten.detach %2486 : !torch.tensor -> !torch.tensor
    %2490 = torch.aten.detach %2489 : !torch.tensor -> !torch.tensor
    %2491 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2492 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2493 = torch.aten._reshape_alias %2481, %2491, %2492 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2494 = torch.aten.t %546 : !torch.tensor -> !torch.tensor
    %2495 = torch.aten.mm %2493, %2494 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2496 = torch.aten.t %2493 : !torch.tensor -> !torch.tensor
    %2497 = torch.aten.mm %2496, %548 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2498 = torch.aten.t %2497 : !torch.tensor -> !torch.tensor
    %2499 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2500 = torch.aten.sum.dim_IntList %2493, %2499, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2501 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2502 = torch.aten.view %2500, %2501 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2503 = torch.aten.detach %2502 : !torch.tensor -> !torch.tensor
    %2504 = torch.aten.detach %2503 : !torch.tensor -> !torch.tensor
    %2505 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2506 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2507 = torch.aten._reshape_alias %2495, %2505, %2506 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2508 = torch.aten.t %2498 : !torch.tensor -> !torch.tensor
    %2509 = torch.aten.detach %2508 : !torch.tensor -> !torch.tensor
    %2510 = torch.aten.detach %2509 : !torch.tensor -> !torch.tensor
    %2511 = torch.aten.gelu_backward %2507, %544, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2512 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %2513 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2514 = torch.aten._reshape_alias %2511, %2512, %2513 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2515 = torch.aten.t %539 : !torch.tensor -> !torch.tensor
    %2516 = torch.aten.mm %2514, %2515 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2517 = torch.aten.t %2514 : !torch.tensor -> !torch.tensor
    %2518 = torch.aten.mm %2517, %541 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2519 = torch.aten.t %2518 : !torch.tensor -> !torch.tensor
    %2520 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2521 = torch.aten.sum.dim_IntList %2514, %2520, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2522 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %2523 = torch.aten.view %2521, %2522 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2524 = torch.aten.detach %2523 : !torch.tensor -> !torch.tensor
    %2525 = torch.aten.detach %2524 : !torch.tensor -> !torch.tensor
    %2526 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2527 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2528 = torch.aten._reshape_alias %2516, %2526, %2527 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2529 = torch.aten.add.Tensor %2481, %2528, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2530 = torch.aten.t %2519 : !torch.tensor -> !torch.tensor
    %2531 = torch.aten.detach %2530 : !torch.tensor -> !torch.tensor
    %2532 = torch.aten.detach %2531 : !torch.tensor -> !torch.tensor
    %2533 = torch.aten.sub.Tensor %537, %result1_34, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2534 = torch.aten.mul.Tensor %2533, %result2_35 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2535 = torch.aten.mul.Tensor %2529, %arg119 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2536 = torch.aten.mul.Scalar %2535, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2537 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2538 = torch.aten.sum.dim_IntList %2535, %2537, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2539 = torch.aten.mul.Tensor %2535, %2534 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2540 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2541 = torch.aten.sum.dim_IntList %2539, %2540, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2542 = torch.aten.mul.Tensor %2534, %2541 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2543 = torch.aten.sub.Tensor %2536, %2538, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2544 = torch.aten.sub.Tensor %2543, %2542, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2545 = torch.aten.div.Scalar %result2_35, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2546 = torch.aten.mul.Tensor %2545, %2544 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2547 = torch.aten.mul.Tensor %2529, %2534 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2548 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2549 = torch.aten.sum.dim_IntList %2547, %2548, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2550 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2551 = torch.aten.sum.dim_IntList %2529, %2550, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2552 = torch.aten.detach %2549 : !torch.tensor -> !torch.tensor
    %2553 = torch.aten.detach %2552 : !torch.tensor -> !torch.tensor
    %2554 = torch.aten.detach %2551 : !torch.tensor -> !torch.tensor
    %2555 = torch.aten.detach %2554 : !torch.tensor -> !torch.tensor
    %2556 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2557 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2558 = torch.aten._reshape_alias %2546, %2556, %2557 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2559 = torch.aten.t %531 : !torch.tensor -> !torch.tensor
    %2560 = torch.aten.mm %2558, %2559 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2561 = torch.aten.t %2558 : !torch.tensor -> !torch.tensor
    %2562 = torch.aten.mm %2561, %533 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2563 = torch.aten.t %2562 : !torch.tensor -> !torch.tensor
    %2564 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2565 = torch.aten.sum.dim_IntList %2558, %2564, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2566 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2567 = torch.aten.view %2565, %2566 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2568 = torch.aten.detach %2567 : !torch.tensor -> !torch.tensor
    %2569 = torch.aten.detach %2568 : !torch.tensor -> !torch.tensor
    %2570 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2571 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2572 = torch.aten._reshape_alias %2560, %2570, %2571 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2573 = torch.aten.t %2563 : !torch.tensor -> !torch.tensor
    %2574 = torch.aten.detach %2573 : !torch.tensor -> !torch.tensor
    %2575 = torch.aten.detach %2574 : !torch.tensor -> !torch.tensor
    %2576 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2577 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2578 = torch.aten._reshape_alias %2572, %2576, %2577 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2579 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2580 = torch.aten.permute %2578, %2579 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2581 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2582 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2583 = torch.aten._reshape_alias %2580, %2581, %2582 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2584 = torch.aten.transpose.int %517, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2585 = torch.aten.bmm %2584, %2583 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2586 = torch.aten.transpose.int %522, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2587 = torch.aten.bmm %2583, %2586 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2588 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2589 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2590 = torch.aten._reshape_alias %2585, %2588, %2589 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2591 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2592 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2593 = torch.aten._reshape_alias %2587, %2591, %2592 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2594 = torch.aten.detach %512 : !torch.tensor -> !torch.tensor
    %2595 = torch.aten._softmax_backward_data %2593, %2594, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2596 = torch.aten.div.Scalar %2595, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %2597 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2598 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2599 = torch.aten._reshape_alias %2596, %2597, %2598 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2600 = torch.aten.transpose.int %499, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2601 = torch.aten.bmm %2600, %2599 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2602 = torch.aten.transpose.int %504, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2603 = torch.aten.bmm %2599, %2602 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2604 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2605 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2606 = torch.aten._reshape_alias %2601, %2604, %2605 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2607 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2608 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2609 = torch.aten._reshape_alias %2603, %2607, %2608 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2610 = torch.aten.transpose.int %2606, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2611 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2612 = torch.aten.permute %2609, %2611 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2613 = torch.aten.clone %2612, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2614 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2615 = torch.aten._unsafe_view %2613, %2614 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2616 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2617 = torch.aten.permute %2590, %2616 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2618 = torch.aten.clone %2617, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2619 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2620 = torch.aten._unsafe_view %2618, %2619 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2621 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2622 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2623 = torch.aten._reshape_alias %2620, %2621, %2622 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2624 = torch.aten.t %480 : !torch.tensor -> !torch.tensor
    %2625 = torch.aten.mm %2623, %2624 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2626 = torch.aten.t %2623 : !torch.tensor -> !torch.tensor
    %2627 = torch.aten.mm %2626, %482 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2628 = torch.aten.t %2627 : !torch.tensor -> !torch.tensor
    %2629 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2630 = torch.aten.sum.dim_IntList %2623, %2629, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2631 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2632 = torch.aten.view %2630, %2631 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2633 = torch.aten.detach %2632 : !torch.tensor -> !torch.tensor
    %2634 = torch.aten.detach %2633 : !torch.tensor -> !torch.tensor
    %2635 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2636 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2637 = torch.aten._reshape_alias %2625, %2635, %2636 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2638 = torch.aten.add.Tensor %2546, %2637, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2639 = torch.aten.t %2628 : !torch.tensor -> !torch.tensor
    %2640 = torch.aten.detach %2639 : !torch.tensor -> !torch.tensor
    %2641 = torch.aten.detach %2640 : !torch.tensor -> !torch.tensor
    %2642 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2643 = torch.aten.permute %2610, %2642 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2644 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2645 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2646 = torch.aten._reshape_alias %2643, %2644, %2645 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2647 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2648 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %2649 = torch.aten._reshape_alias %2646, %2647, %2648 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2650 = torch.aten.t %470 : !torch.tensor -> !torch.tensor
    %2651 = torch.aten.mm %2649, %2650 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2652 = torch.aten.t %2649 : !torch.tensor -> !torch.tensor
    %2653 = torch.aten.mm %2652, %472 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2654 = torch.aten.t %2653 : !torch.tensor -> !torch.tensor
    %2655 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2656 = torch.aten.sum.dim_IntList %2649, %2655, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2657 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2658 = torch.aten.view %2656, %2657 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2659 = torch.aten.detach %2658 : !torch.tensor -> !torch.tensor
    %2660 = torch.aten.detach %2659 : !torch.tensor -> !torch.tensor
    %2661 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2662 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2663 = torch.aten._reshape_alias %2651, %2661, %2662 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2664 = torch.aten.add.Tensor %2638, %2663, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2665 = torch.aten.t %2654 : !torch.tensor -> !torch.tensor
    %2666 = torch.aten.detach %2665 : !torch.tensor -> !torch.tensor
    %2667 = torch.aten.detach %2666 : !torch.tensor -> !torch.tensor
    %2668 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2669 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2670 = torch.aten._reshape_alias %2615, %2668, %2669 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2671 = torch.aten.t %464 : !torch.tensor -> !torch.tensor
    %2672 = torch.aten.mm %2670, %2671 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2673 = torch.aten.t %2670 : !torch.tensor -> !torch.tensor
    %2674 = torch.aten.mm %2673, %466 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2675 = torch.aten.t %2674 : !torch.tensor -> !torch.tensor
    %2676 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2677 = torch.aten.sum.dim_IntList %2670, %2676, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2678 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2679 = torch.aten.view %2677, %2678 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2680 = torch.aten.detach %2679 : !torch.tensor -> !torch.tensor
    %2681 = torch.aten.detach %2680 : !torch.tensor -> !torch.tensor
    %2682 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2683 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2684 = torch.aten._reshape_alias %2672, %2682, %2683 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2685 = torch.aten.add.Tensor %2664, %2684, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2686 = torch.aten.t %2675 : !torch.tensor -> !torch.tensor
    %2687 = torch.aten.detach %2686 : !torch.tensor -> !torch.tensor
    %2688 = torch.aten.detach %2687 : !torch.tensor -> !torch.tensor
    %2689 = torch.aten.sub.Tensor %462, %result1_31, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2690 = torch.aten.mul.Tensor %2689, %result2_32 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2691 = torch.aten.mul.Tensor %2685, %arg115 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2692 = torch.aten.mul.Scalar %2691, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2693 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2694 = torch.aten.sum.dim_IntList %2691, %2693, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2695 = torch.aten.mul.Tensor %2691, %2690 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2696 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2697 = torch.aten.sum.dim_IntList %2695, %2696, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2698 = torch.aten.mul.Tensor %2690, %2697 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2699 = torch.aten.sub.Tensor %2692, %2694, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2700 = torch.aten.sub.Tensor %2699, %2698, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2701 = torch.aten.div.Scalar %result2_32, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2702 = torch.aten.mul.Tensor %2701, %2700 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2703 = torch.aten.mul.Tensor %2685, %2690 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2704 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2705 = torch.aten.sum.dim_IntList %2703, %2704, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2706 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2707 = torch.aten.sum.dim_IntList %2685, %2706, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2708 = torch.aten.detach %2705 : !torch.tensor -> !torch.tensor
    %2709 = torch.aten.detach %2708 : !torch.tensor -> !torch.tensor
    %2710 = torch.aten.detach %2707 : !torch.tensor -> !torch.tensor
    %2711 = torch.aten.detach %2710 : !torch.tensor -> !torch.tensor
    %2712 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2713 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2714 = torch.aten._reshape_alias %2702, %2712, %2713 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2715 = torch.aten.t %456 : !torch.tensor -> !torch.tensor
    %2716 = torch.aten.mm %2714, %2715 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2717 = torch.aten.t %2714 : !torch.tensor -> !torch.tensor
    %2718 = torch.aten.mm %2717, %458 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2719 = torch.aten.t %2718 : !torch.tensor -> !torch.tensor
    %2720 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2721 = torch.aten.sum.dim_IntList %2714, %2720, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2722 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2723 = torch.aten.view %2721, %2722 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2724 = torch.aten.detach %2723 : !torch.tensor -> !torch.tensor
    %2725 = torch.aten.detach %2724 : !torch.tensor -> !torch.tensor
    %2726 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2727 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2728 = torch.aten._reshape_alias %2716, %2726, %2727 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2729 = torch.aten.t %2719 : !torch.tensor -> !torch.tensor
    %2730 = torch.aten.detach %2729 : !torch.tensor -> !torch.tensor
    %2731 = torch.aten.detach %2730 : !torch.tensor -> !torch.tensor
    %2732 = torch.aten.gelu_backward %2728, %454, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2733 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %2734 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2735 = torch.aten._reshape_alias %2732, %2733, %2734 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2736 = torch.aten.t %449 : !torch.tensor -> !torch.tensor
    %2737 = torch.aten.mm %2735, %2736 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2738 = torch.aten.t %2735 : !torch.tensor -> !torch.tensor
    %2739 = torch.aten.mm %2738, %451 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2740 = torch.aten.t %2739 : !torch.tensor -> !torch.tensor
    %2741 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2742 = torch.aten.sum.dim_IntList %2735, %2741, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2743 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %2744 = torch.aten.view %2742, %2743 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2745 = torch.aten.detach %2744 : !torch.tensor -> !torch.tensor
    %2746 = torch.aten.detach %2745 : !torch.tensor -> !torch.tensor
    %2747 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2748 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2749 = torch.aten._reshape_alias %2737, %2747, %2748 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2750 = torch.aten.add.Tensor %2702, %2749, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2751 = torch.aten.t %2740 : !torch.tensor -> !torch.tensor
    %2752 = torch.aten.detach %2751 : !torch.tensor -> !torch.tensor
    %2753 = torch.aten.detach %2752 : !torch.tensor -> !torch.tensor
    %2754 = torch.aten.sub.Tensor %447, %result1_28, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2755 = torch.aten.mul.Tensor %2754, %result2_29 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2756 = torch.aten.mul.Tensor %2750, %arg103 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2757 = torch.aten.mul.Scalar %2756, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2758 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2759 = torch.aten.sum.dim_IntList %2756, %2758, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2760 = torch.aten.mul.Tensor %2756, %2755 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2761 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2762 = torch.aten.sum.dim_IntList %2760, %2761, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2763 = torch.aten.mul.Tensor %2755, %2762 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2764 = torch.aten.sub.Tensor %2757, %2759, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2765 = torch.aten.sub.Tensor %2764, %2763, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2766 = torch.aten.div.Scalar %result2_29, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2767 = torch.aten.mul.Tensor %2766, %2765 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2768 = torch.aten.mul.Tensor %2750, %2755 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2769 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2770 = torch.aten.sum.dim_IntList %2768, %2769, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2771 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2772 = torch.aten.sum.dim_IntList %2750, %2771, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2773 = torch.aten.detach %2770 : !torch.tensor -> !torch.tensor
    %2774 = torch.aten.detach %2773 : !torch.tensor -> !torch.tensor
    %2775 = torch.aten.detach %2772 : !torch.tensor -> !torch.tensor
    %2776 = torch.aten.detach %2775 : !torch.tensor -> !torch.tensor
    %2777 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2778 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2779 = torch.aten._reshape_alias %2767, %2777, %2778 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2780 = torch.aten.t %441 : !torch.tensor -> !torch.tensor
    %2781 = torch.aten.mm %2779, %2780 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2782 = torch.aten.t %2779 : !torch.tensor -> !torch.tensor
    %2783 = torch.aten.mm %2782, %443 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2784 = torch.aten.t %2783 : !torch.tensor -> !torch.tensor
    %2785 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2786 = torch.aten.sum.dim_IntList %2779, %2785, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2787 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2788 = torch.aten.view %2786, %2787 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2789 = torch.aten.detach %2788 : !torch.tensor -> !torch.tensor
    %2790 = torch.aten.detach %2789 : !torch.tensor -> !torch.tensor
    %2791 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2792 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2793 = torch.aten._reshape_alias %2781, %2791, %2792 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2794 = torch.aten.t %2784 : !torch.tensor -> !torch.tensor
    %2795 = torch.aten.detach %2794 : !torch.tensor -> !torch.tensor
    %2796 = torch.aten.detach %2795 : !torch.tensor -> !torch.tensor
    %2797 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2798 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2799 = torch.aten._reshape_alias %2793, %2797, %2798 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2800 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2801 = torch.aten.permute %2799, %2800 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2802 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2803 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2804 = torch.aten._reshape_alias %2801, %2802, %2803 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2805 = torch.aten.transpose.int %427, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2806 = torch.aten.bmm %2805, %2804 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2807 = torch.aten.transpose.int %432, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2808 = torch.aten.bmm %2804, %2807 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2809 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2810 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2811 = torch.aten._reshape_alias %2806, %2809, %2810 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2812 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2813 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2814 = torch.aten._reshape_alias %2808, %2812, %2813 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2815 = torch.aten.detach %422 : !torch.tensor -> !torch.tensor
    %2816 = torch.aten._softmax_backward_data %2814, %2815, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2817 = torch.aten.div.Scalar %2816, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %2818 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2819 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2820 = torch.aten._reshape_alias %2817, %2818, %2819 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2821 = torch.aten.transpose.int %409, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2822 = torch.aten.bmm %2821, %2820 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2823 = torch.aten.transpose.int %414, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2824 = torch.aten.bmm %2820, %2823 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2825 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2826 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2827 = torch.aten._reshape_alias %2822, %2825, %2826 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2828 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2829 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2830 = torch.aten._reshape_alias %2824, %2828, %2829 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2831 = torch.aten.transpose.int %2827, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2832 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2833 = torch.aten.permute %2830, %2832 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2834 = torch.aten.clone %2833, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2835 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2836 = torch.aten._unsafe_view %2834, %2835 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2837 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2838 = torch.aten.permute %2811, %2837 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2839 = torch.aten.clone %2838, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2840 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2841 = torch.aten._unsafe_view %2839, %2840 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2842 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2843 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2844 = torch.aten._reshape_alias %2841, %2842, %2843 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2845 = torch.aten.t %390 : !torch.tensor -> !torch.tensor
    %2846 = torch.aten.mm %2844, %2845 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2847 = torch.aten.t %2844 : !torch.tensor -> !torch.tensor
    %2848 = torch.aten.mm %2847, %392 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2849 = torch.aten.t %2848 : !torch.tensor -> !torch.tensor
    %2850 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2851 = torch.aten.sum.dim_IntList %2844, %2850, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2852 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2853 = torch.aten.view %2851, %2852 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2854 = torch.aten.detach %2853 : !torch.tensor -> !torch.tensor
    %2855 = torch.aten.detach %2854 : !torch.tensor -> !torch.tensor
    %2856 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2857 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2858 = torch.aten._reshape_alias %2846, %2856, %2857 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2859 = torch.aten.add.Tensor %2767, %2858, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2860 = torch.aten.t %2849 : !torch.tensor -> !torch.tensor
    %2861 = torch.aten.detach %2860 : !torch.tensor -> !torch.tensor
    %2862 = torch.aten.detach %2861 : !torch.tensor -> !torch.tensor
    %2863 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2864 = torch.aten.permute %2831, %2863 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2865 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2866 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2867 = torch.aten._reshape_alias %2864, %2865, %2866 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2868 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2869 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %2870 = torch.aten._reshape_alias %2867, %2868, %2869 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2871 = torch.aten.t %380 : !torch.tensor -> !torch.tensor
    %2872 = torch.aten.mm %2870, %2871 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2873 = torch.aten.t %2870 : !torch.tensor -> !torch.tensor
    %2874 = torch.aten.mm %2873, %382 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2875 = torch.aten.t %2874 : !torch.tensor -> !torch.tensor
    %2876 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2877 = torch.aten.sum.dim_IntList %2870, %2876, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2878 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2879 = torch.aten.view %2877, %2878 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2880 = torch.aten.detach %2879 : !torch.tensor -> !torch.tensor
    %2881 = torch.aten.detach %2880 : !torch.tensor -> !torch.tensor
    %2882 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2883 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2884 = torch.aten._reshape_alias %2872, %2882, %2883 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2885 = torch.aten.add.Tensor %2859, %2884, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2886 = torch.aten.t %2875 : !torch.tensor -> !torch.tensor
    %2887 = torch.aten.detach %2886 : !torch.tensor -> !torch.tensor
    %2888 = torch.aten.detach %2887 : !torch.tensor -> !torch.tensor
    %2889 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2890 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2891 = torch.aten._reshape_alias %2836, %2889, %2890 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2892 = torch.aten.t %374 : !torch.tensor -> !torch.tensor
    %2893 = torch.aten.mm %2891, %2892 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2894 = torch.aten.t %2891 : !torch.tensor -> !torch.tensor
    %2895 = torch.aten.mm %2894, %376 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2896 = torch.aten.t %2895 : !torch.tensor -> !torch.tensor
    %2897 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2898 = torch.aten.sum.dim_IntList %2891, %2897, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2899 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2900 = torch.aten.view %2898, %2899 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2901 = torch.aten.detach %2900 : !torch.tensor -> !torch.tensor
    %2902 = torch.aten.detach %2901 : !torch.tensor -> !torch.tensor
    %2903 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2904 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2905 = torch.aten._reshape_alias %2893, %2903, %2904 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2906 = torch.aten.add.Tensor %2885, %2905, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2907 = torch.aten.t %2896 : !torch.tensor -> !torch.tensor
    %2908 = torch.aten.detach %2907 : !torch.tensor -> !torch.tensor
    %2909 = torch.aten.detach %2908 : !torch.tensor -> !torch.tensor
    %2910 = torch.aten.sub.Tensor %372, %result1_25, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2911 = torch.aten.mul.Tensor %2910, %result2_26 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2912 = torch.aten.mul.Tensor %2906, %arg99 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2913 = torch.aten.mul.Scalar %2912, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2914 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2915 = torch.aten.sum.dim_IntList %2912, %2914, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2916 = torch.aten.mul.Tensor %2912, %2911 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2917 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2918 = torch.aten.sum.dim_IntList %2916, %2917, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2919 = torch.aten.mul.Tensor %2911, %2918 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2920 = torch.aten.sub.Tensor %2913, %2915, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2921 = torch.aten.sub.Tensor %2920, %2919, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2922 = torch.aten.div.Scalar %result2_26, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2923 = torch.aten.mul.Tensor %2922, %2921 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2924 = torch.aten.mul.Tensor %2906, %2911 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2925 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2926 = torch.aten.sum.dim_IntList %2924, %2925, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2927 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2928 = torch.aten.sum.dim_IntList %2906, %2927, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2929 = torch.aten.detach %2926 : !torch.tensor -> !torch.tensor
    %2930 = torch.aten.detach %2929 : !torch.tensor -> !torch.tensor
    %2931 = torch.aten.detach %2928 : !torch.tensor -> !torch.tensor
    %2932 = torch.aten.detach %2931 : !torch.tensor -> !torch.tensor
    %2933 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2934 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2935 = torch.aten._reshape_alias %2923, %2933, %2934 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2936 = torch.aten.t %366 : !torch.tensor -> !torch.tensor
    %2937 = torch.aten.mm %2935, %2936 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2938 = torch.aten.t %2935 : !torch.tensor -> !torch.tensor
    %2939 = torch.aten.mm %2938, %368 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2940 = torch.aten.t %2939 : !torch.tensor -> !torch.tensor
    %2941 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2942 = torch.aten.sum.dim_IntList %2935, %2941, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2943 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %2944 = torch.aten.view %2942, %2943 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2945 = torch.aten.detach %2944 : !torch.tensor -> !torch.tensor
    %2946 = torch.aten.detach %2945 : !torch.tensor -> !torch.tensor
    %2947 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2948 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2949 = torch.aten._reshape_alias %2937, %2947, %2948 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2950 = torch.aten.t %2940 : !torch.tensor -> !torch.tensor
    %2951 = torch.aten.detach %2950 : !torch.tensor -> !torch.tensor
    %2952 = torch.aten.detach %2951 : !torch.tensor -> !torch.tensor
    %2953 = torch.aten.gelu_backward %2949, %364, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2954 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %2955 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2956 = torch.aten._reshape_alias %2953, %2954, %2955 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2957 = torch.aten.t %359 : !torch.tensor -> !torch.tensor
    %2958 = torch.aten.mm %2956, %2957 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2959 = torch.aten.t %2956 : !torch.tensor -> !torch.tensor
    %2960 = torch.aten.mm %2959, %361 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2961 = torch.aten.t %2960 : !torch.tensor -> !torch.tensor
    %2962 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2963 = torch.aten.sum.dim_IntList %2956, %2962, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2964 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %2965 = torch.aten.view %2963, %2964 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2966 = torch.aten.detach %2965 : !torch.tensor -> !torch.tensor
    %2967 = torch.aten.detach %2966 : !torch.tensor -> !torch.tensor
    %2968 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2969 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2970 = torch.aten._reshape_alias %2958, %2968, %2969 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %2971 = torch.aten.add.Tensor %2923, %2970, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2972 = torch.aten.t %2961 : !torch.tensor -> !torch.tensor
    %2973 = torch.aten.detach %2972 : !torch.tensor -> !torch.tensor
    %2974 = torch.aten.detach %2973 : !torch.tensor -> !torch.tensor
    %2975 = torch.aten.sub.Tensor %357, %result1_22, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2976 = torch.aten.mul.Tensor %2975, %result2_23 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2977 = torch.aten.mul.Tensor %2971, %arg87 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2978 = torch.aten.mul.Scalar %2977, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2979 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2980 = torch.aten.sum.dim_IntList %2977, %2979, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2981 = torch.aten.mul.Tensor %2977, %2976 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2982 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2983 = torch.aten.sum.dim_IntList %2981, %2982, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2984 = torch.aten.mul.Tensor %2976, %2983 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2985 = torch.aten.sub.Tensor %2978, %2980, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2986 = torch.aten.sub.Tensor %2985, %2984, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2987 = torch.aten.div.Scalar %result2_23, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %2988 = torch.aten.mul.Tensor %2987, %2986 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2989 = torch.aten.mul.Tensor %2971, %2976 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2990 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2991 = torch.aten.sum.dim_IntList %2989, %2990, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2992 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2993 = torch.aten.sum.dim_IntList %2971, %2992, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2994 = torch.aten.detach %2991 : !torch.tensor -> !torch.tensor
    %2995 = torch.aten.detach %2994 : !torch.tensor -> !torch.tensor
    %2996 = torch.aten.detach %2993 : !torch.tensor -> !torch.tensor
    %2997 = torch.aten.detach %2996 : !torch.tensor -> !torch.tensor
    %2998 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %2999 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3000 = torch.aten._reshape_alias %2988, %2998, %2999 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3001 = torch.aten.t %351 : !torch.tensor -> !torch.tensor
    %3002 = torch.aten.mm %3000, %3001 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3003 = torch.aten.t %3000 : !torch.tensor -> !torch.tensor
    %3004 = torch.aten.mm %3003, %353 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3005 = torch.aten.t %3004 : !torch.tensor -> !torch.tensor
    %3006 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3007 = torch.aten.sum.dim_IntList %3000, %3006, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3008 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3009 = torch.aten.view %3007, %3008 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3010 = torch.aten.detach %3009 : !torch.tensor -> !torch.tensor
    %3011 = torch.aten.detach %3010 : !torch.tensor -> !torch.tensor
    %3012 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3013 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3014 = torch.aten._reshape_alias %3002, %3012, %3013 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3015 = torch.aten.t %3005 : !torch.tensor -> !torch.tensor
    %3016 = torch.aten.detach %3015 : !torch.tensor -> !torch.tensor
    %3017 = torch.aten.detach %3016 : !torch.tensor -> !torch.tensor
    %3018 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3019 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3020 = torch.aten._reshape_alias %3014, %3018, %3019 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3021 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3022 = torch.aten.permute %3020, %3021 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3023 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3024 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3025 = torch.aten._reshape_alias %3022, %3023, %3024 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3026 = torch.aten.transpose.int %337, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3027 = torch.aten.bmm %3026, %3025 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3028 = torch.aten.transpose.int %342, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3029 = torch.aten.bmm %3025, %3028 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3030 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3031 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3032 = torch.aten._reshape_alias %3027, %3030, %3031 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3033 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3034 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3035 = torch.aten._reshape_alias %3029, %3033, %3034 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3036 = torch.aten.detach %332 : !torch.tensor -> !torch.tensor
    %3037 = torch.aten._softmax_backward_data %3035, %3036, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3038 = torch.aten.div.Scalar %3037, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %3039 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3040 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3041 = torch.aten._reshape_alias %3038, %3039, %3040 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3042 = torch.aten.transpose.int %319, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3043 = torch.aten.bmm %3042, %3041 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3044 = torch.aten.transpose.int %324, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3045 = torch.aten.bmm %3041, %3044 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3046 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3047 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3048 = torch.aten._reshape_alias %3043, %3046, %3047 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3049 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3050 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3051 = torch.aten._reshape_alias %3045, %3049, %3050 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3052 = torch.aten.transpose.int %3048, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3053 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3054 = torch.aten.permute %3051, %3053 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3055 = torch.aten.clone %3054, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3056 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3057 = torch.aten._unsafe_view %3055, %3056 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3058 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3059 = torch.aten.permute %3032, %3058 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3060 = torch.aten.clone %3059, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3061 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3062 = torch.aten._unsafe_view %3060, %3061 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3063 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3064 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3065 = torch.aten._reshape_alias %3062, %3063, %3064 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3066 = torch.aten.t %300 : !torch.tensor -> !torch.tensor
    %3067 = torch.aten.mm %3065, %3066 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3068 = torch.aten.t %3065 : !torch.tensor -> !torch.tensor
    %3069 = torch.aten.mm %3068, %302 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3070 = torch.aten.t %3069 : !torch.tensor -> !torch.tensor
    %3071 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3072 = torch.aten.sum.dim_IntList %3065, %3071, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3073 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3074 = torch.aten.view %3072, %3073 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3075 = torch.aten.detach %3074 : !torch.tensor -> !torch.tensor
    %3076 = torch.aten.detach %3075 : !torch.tensor -> !torch.tensor
    %3077 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3078 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3079 = torch.aten._reshape_alias %3067, %3077, %3078 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3080 = torch.aten.add.Tensor %2988, %3079, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3081 = torch.aten.t %3070 : !torch.tensor -> !torch.tensor
    %3082 = torch.aten.detach %3081 : !torch.tensor -> !torch.tensor
    %3083 = torch.aten.detach %3082 : !torch.tensor -> !torch.tensor
    %3084 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3085 = torch.aten.permute %3052, %3084 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3086 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3087 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3088 = torch.aten._reshape_alias %3085, %3086, %3087 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3089 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3090 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %3091 = torch.aten._reshape_alias %3088, %3089, %3090 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3092 = torch.aten.t %290 : !torch.tensor -> !torch.tensor
    %3093 = torch.aten.mm %3091, %3092 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3094 = torch.aten.t %3091 : !torch.tensor -> !torch.tensor
    %3095 = torch.aten.mm %3094, %292 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3096 = torch.aten.t %3095 : !torch.tensor -> !torch.tensor
    %3097 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3098 = torch.aten.sum.dim_IntList %3091, %3097, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3099 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3100 = torch.aten.view %3098, %3099 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3101 = torch.aten.detach %3100 : !torch.tensor -> !torch.tensor
    %3102 = torch.aten.detach %3101 : !torch.tensor -> !torch.tensor
    %3103 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3104 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3105 = torch.aten._reshape_alias %3093, %3103, %3104 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3106 = torch.aten.add.Tensor %3080, %3105, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3107 = torch.aten.t %3096 : !torch.tensor -> !torch.tensor
    %3108 = torch.aten.detach %3107 : !torch.tensor -> !torch.tensor
    %3109 = torch.aten.detach %3108 : !torch.tensor -> !torch.tensor
    %3110 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3111 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3112 = torch.aten._reshape_alias %3057, %3110, %3111 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3113 = torch.aten.t %284 : !torch.tensor -> !torch.tensor
    %3114 = torch.aten.mm %3112, %3113 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3115 = torch.aten.t %3112 : !torch.tensor -> !torch.tensor
    %3116 = torch.aten.mm %3115, %286 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3117 = torch.aten.t %3116 : !torch.tensor -> !torch.tensor
    %3118 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3119 = torch.aten.sum.dim_IntList %3112, %3118, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3120 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3121 = torch.aten.view %3119, %3120 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3122 = torch.aten.detach %3121 : !torch.tensor -> !torch.tensor
    %3123 = torch.aten.detach %3122 : !torch.tensor -> !torch.tensor
    %3124 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3125 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3126 = torch.aten._reshape_alias %3114, %3124, %3125 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3127 = torch.aten.add.Tensor %3106, %3126, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3128 = torch.aten.t %3117 : !torch.tensor -> !torch.tensor
    %3129 = torch.aten.detach %3128 : !torch.tensor -> !torch.tensor
    %3130 = torch.aten.detach %3129 : !torch.tensor -> !torch.tensor
    %3131 = torch.aten.sub.Tensor %282, %result1_19, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3132 = torch.aten.mul.Tensor %3131, %result2_20 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3133 = torch.aten.mul.Tensor %3127, %arg83 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3134 = torch.aten.mul.Scalar %3133, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3135 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3136 = torch.aten.sum.dim_IntList %3133, %3135, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3137 = torch.aten.mul.Tensor %3133, %3132 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3138 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3139 = torch.aten.sum.dim_IntList %3137, %3138, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3140 = torch.aten.mul.Tensor %3132, %3139 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3141 = torch.aten.sub.Tensor %3134, %3136, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3142 = torch.aten.sub.Tensor %3141, %3140, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3143 = torch.aten.div.Scalar %result2_20, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3144 = torch.aten.mul.Tensor %3143, %3142 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3145 = torch.aten.mul.Tensor %3127, %3132 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3146 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3147 = torch.aten.sum.dim_IntList %3145, %3146, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3148 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3149 = torch.aten.sum.dim_IntList %3127, %3148, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3150 = torch.aten.detach %3147 : !torch.tensor -> !torch.tensor
    %3151 = torch.aten.detach %3150 : !torch.tensor -> !torch.tensor
    %3152 = torch.aten.detach %3149 : !torch.tensor -> !torch.tensor
    %3153 = torch.aten.detach %3152 : !torch.tensor -> !torch.tensor
    %3154 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3155 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3156 = torch.aten._reshape_alias %3144, %3154, %3155 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3157 = torch.aten.t %276 : !torch.tensor -> !torch.tensor
    %3158 = torch.aten.mm %3156, %3157 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3159 = torch.aten.t %3156 : !torch.tensor -> !torch.tensor
    %3160 = torch.aten.mm %3159, %278 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3161 = torch.aten.t %3160 : !torch.tensor -> !torch.tensor
    %3162 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3163 = torch.aten.sum.dim_IntList %3156, %3162, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3164 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3165 = torch.aten.view %3163, %3164 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3166 = torch.aten.detach %3165 : !torch.tensor -> !torch.tensor
    %3167 = torch.aten.detach %3166 : !torch.tensor -> !torch.tensor
    %3168 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3169 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3170 = torch.aten._reshape_alias %3158, %3168, %3169 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3171 = torch.aten.t %3161 : !torch.tensor -> !torch.tensor
    %3172 = torch.aten.detach %3171 : !torch.tensor -> !torch.tensor
    %3173 = torch.aten.detach %3172 : !torch.tensor -> !torch.tensor
    %3174 = torch.aten.gelu_backward %3170, %274, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %3175 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %3176 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3177 = torch.aten._reshape_alias %3174, %3175, %3176 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3178 = torch.aten.t %269 : !torch.tensor -> !torch.tensor
    %3179 = torch.aten.mm %3177, %3178 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3180 = torch.aten.t %3177 : !torch.tensor -> !torch.tensor
    %3181 = torch.aten.mm %3180, %271 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3182 = torch.aten.t %3181 : !torch.tensor -> !torch.tensor
    %3183 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3184 = torch.aten.sum.dim_IntList %3177, %3183, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3185 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %3186 = torch.aten.view %3184, %3185 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3187 = torch.aten.detach %3186 : !torch.tensor -> !torch.tensor
    %3188 = torch.aten.detach %3187 : !torch.tensor -> !torch.tensor
    %3189 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3190 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3191 = torch.aten._reshape_alias %3179, %3189, %3190 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3192 = torch.aten.add.Tensor %3144, %3191, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3193 = torch.aten.t %3182 : !torch.tensor -> !torch.tensor
    %3194 = torch.aten.detach %3193 : !torch.tensor -> !torch.tensor
    %3195 = torch.aten.detach %3194 : !torch.tensor -> !torch.tensor
    %3196 = torch.aten.sub.Tensor %267, %result1_16, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3197 = torch.aten.mul.Tensor %3196, %result2_17 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3198 = torch.aten.mul.Tensor %3192, %arg71 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3199 = torch.aten.mul.Scalar %3198, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3200 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3201 = torch.aten.sum.dim_IntList %3198, %3200, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3202 = torch.aten.mul.Tensor %3198, %3197 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3203 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3204 = torch.aten.sum.dim_IntList %3202, %3203, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3205 = torch.aten.mul.Tensor %3197, %3204 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3206 = torch.aten.sub.Tensor %3199, %3201, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3207 = torch.aten.sub.Tensor %3206, %3205, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3208 = torch.aten.div.Scalar %result2_17, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3209 = torch.aten.mul.Tensor %3208, %3207 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3210 = torch.aten.mul.Tensor %3192, %3197 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3211 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3212 = torch.aten.sum.dim_IntList %3210, %3211, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3213 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3214 = torch.aten.sum.dim_IntList %3192, %3213, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3215 = torch.aten.detach %3212 : !torch.tensor -> !torch.tensor
    %3216 = torch.aten.detach %3215 : !torch.tensor -> !torch.tensor
    %3217 = torch.aten.detach %3214 : !torch.tensor -> !torch.tensor
    %3218 = torch.aten.detach %3217 : !torch.tensor -> !torch.tensor
    %3219 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3220 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3221 = torch.aten._reshape_alias %3209, %3219, %3220 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3222 = torch.aten.t %261 : !torch.tensor -> !torch.tensor
    %3223 = torch.aten.mm %3221, %3222 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3224 = torch.aten.t %3221 : !torch.tensor -> !torch.tensor
    %3225 = torch.aten.mm %3224, %263 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3226 = torch.aten.t %3225 : !torch.tensor -> !torch.tensor
    %3227 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3228 = torch.aten.sum.dim_IntList %3221, %3227, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3229 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3230 = torch.aten.view %3228, %3229 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3231 = torch.aten.detach %3230 : !torch.tensor -> !torch.tensor
    %3232 = torch.aten.detach %3231 : !torch.tensor -> !torch.tensor
    %3233 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3234 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3235 = torch.aten._reshape_alias %3223, %3233, %3234 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3236 = torch.aten.t %3226 : !torch.tensor -> !torch.tensor
    %3237 = torch.aten.detach %3236 : !torch.tensor -> !torch.tensor
    %3238 = torch.aten.detach %3237 : !torch.tensor -> !torch.tensor
    %3239 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3240 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3241 = torch.aten._reshape_alias %3235, %3239, %3240 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3242 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3243 = torch.aten.permute %3241, %3242 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3244 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3245 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3246 = torch.aten._reshape_alias %3243, %3244, %3245 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3247 = torch.aten.transpose.int %247, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3248 = torch.aten.bmm %3247, %3246 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3249 = torch.aten.transpose.int %252, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3250 = torch.aten.bmm %3246, %3249 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3251 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3252 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3253 = torch.aten._reshape_alias %3248, %3251, %3252 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3254 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3255 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3256 = torch.aten._reshape_alias %3250, %3254, %3255 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3257 = torch.aten.detach %242 : !torch.tensor -> !torch.tensor
    %3258 = torch.aten._softmax_backward_data %3256, %3257, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3259 = torch.aten.div.Scalar %3258, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %3260 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3261 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3262 = torch.aten._reshape_alias %3259, %3260, %3261 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3263 = torch.aten.transpose.int %229, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3264 = torch.aten.bmm %3263, %3262 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3265 = torch.aten.transpose.int %234, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3266 = torch.aten.bmm %3262, %3265 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3267 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3268 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3269 = torch.aten._reshape_alias %3264, %3267, %3268 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3270 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3271 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3272 = torch.aten._reshape_alias %3266, %3270, %3271 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3273 = torch.aten.transpose.int %3269, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3274 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3275 = torch.aten.permute %3272, %3274 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3276 = torch.aten.clone %3275, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3277 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3278 = torch.aten._unsafe_view %3276, %3277 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3279 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3280 = torch.aten.permute %3253, %3279 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3281 = torch.aten.clone %3280, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3282 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3283 = torch.aten._unsafe_view %3281, %3282 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3284 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3285 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3286 = torch.aten._reshape_alias %3283, %3284, %3285 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3287 = torch.aten.t %210 : !torch.tensor -> !torch.tensor
    %3288 = torch.aten.mm %3286, %3287 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3289 = torch.aten.t %3286 : !torch.tensor -> !torch.tensor
    %3290 = torch.aten.mm %3289, %212 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3291 = torch.aten.t %3290 : !torch.tensor -> !torch.tensor
    %3292 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3293 = torch.aten.sum.dim_IntList %3286, %3292, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3294 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3295 = torch.aten.view %3293, %3294 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3296 = torch.aten.detach %3295 : !torch.tensor -> !torch.tensor
    %3297 = torch.aten.detach %3296 : !torch.tensor -> !torch.tensor
    %3298 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3299 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3300 = torch.aten._reshape_alias %3288, %3298, %3299 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3301 = torch.aten.add.Tensor %3209, %3300, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3302 = torch.aten.t %3291 : !torch.tensor -> !torch.tensor
    %3303 = torch.aten.detach %3302 : !torch.tensor -> !torch.tensor
    %3304 = torch.aten.detach %3303 : !torch.tensor -> !torch.tensor
    %3305 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3306 = torch.aten.permute %3273, %3305 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3307 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3308 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3309 = torch.aten._reshape_alias %3306, %3307, %3308 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3310 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3311 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %3312 = torch.aten._reshape_alias %3309, %3310, %3311 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3313 = torch.aten.t %200 : !torch.tensor -> !torch.tensor
    %3314 = torch.aten.mm %3312, %3313 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3315 = torch.aten.t %3312 : !torch.tensor -> !torch.tensor
    %3316 = torch.aten.mm %3315, %202 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3317 = torch.aten.t %3316 : !torch.tensor -> !torch.tensor
    %3318 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3319 = torch.aten.sum.dim_IntList %3312, %3318, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3320 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3321 = torch.aten.view %3319, %3320 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3322 = torch.aten.detach %3321 : !torch.tensor -> !torch.tensor
    %3323 = torch.aten.detach %3322 : !torch.tensor -> !torch.tensor
    %3324 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3325 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3326 = torch.aten._reshape_alias %3314, %3324, %3325 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3327 = torch.aten.add.Tensor %3301, %3326, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3328 = torch.aten.t %3317 : !torch.tensor -> !torch.tensor
    %3329 = torch.aten.detach %3328 : !torch.tensor -> !torch.tensor
    %3330 = torch.aten.detach %3329 : !torch.tensor -> !torch.tensor
    %3331 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3332 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3333 = torch.aten._reshape_alias %3278, %3331, %3332 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3334 = torch.aten.t %194 : !torch.tensor -> !torch.tensor
    %3335 = torch.aten.mm %3333, %3334 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3336 = torch.aten.t %3333 : !torch.tensor -> !torch.tensor
    %3337 = torch.aten.mm %3336, %196 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3338 = torch.aten.t %3337 : !torch.tensor -> !torch.tensor
    %3339 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3340 = torch.aten.sum.dim_IntList %3333, %3339, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3341 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3342 = torch.aten.view %3340, %3341 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3343 = torch.aten.detach %3342 : !torch.tensor -> !torch.tensor
    %3344 = torch.aten.detach %3343 : !torch.tensor -> !torch.tensor
    %3345 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3346 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3347 = torch.aten._reshape_alias %3335, %3345, %3346 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3348 = torch.aten.add.Tensor %3327, %3347, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3349 = torch.aten.t %3338 : !torch.tensor -> !torch.tensor
    %3350 = torch.aten.detach %3349 : !torch.tensor -> !torch.tensor
    %3351 = torch.aten.detach %3350 : !torch.tensor -> !torch.tensor
    %3352 = torch.aten.sub.Tensor %192, %result1_13, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3353 = torch.aten.mul.Tensor %3352, %result2_14 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3354 = torch.aten.mul.Tensor %3348, %arg35 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3355 = torch.aten.mul.Scalar %3354, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3356 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3357 = torch.aten.sum.dim_IntList %3354, %3356, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3358 = torch.aten.mul.Tensor %3354, %3353 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3359 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3360 = torch.aten.sum.dim_IntList %3358, %3359, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3361 = torch.aten.mul.Tensor %3353, %3360 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3362 = torch.aten.sub.Tensor %3355, %3357, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3363 = torch.aten.sub.Tensor %3362, %3361, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3364 = torch.aten.div.Scalar %result2_14, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3365 = torch.aten.mul.Tensor %3364, %3363 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3366 = torch.aten.mul.Tensor %3348, %3353 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3367 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3368 = torch.aten.sum.dim_IntList %3366, %3367, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3369 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3370 = torch.aten.sum.dim_IntList %3348, %3369, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3371 = torch.aten.detach %3368 : !torch.tensor -> !torch.tensor
    %3372 = torch.aten.detach %3371 : !torch.tensor -> !torch.tensor
    %3373 = torch.aten.detach %3370 : !torch.tensor -> !torch.tensor
    %3374 = torch.aten.detach %3373 : !torch.tensor -> !torch.tensor
    %3375 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3376 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3377 = torch.aten._reshape_alias %3365, %3375, %3376 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3378 = torch.aten.t %186 : !torch.tensor -> !torch.tensor
    %3379 = torch.aten.mm %3377, %3378 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3380 = torch.aten.t %3377 : !torch.tensor -> !torch.tensor
    %3381 = torch.aten.mm %3380, %188 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3382 = torch.aten.t %3381 : !torch.tensor -> !torch.tensor
    %3383 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3384 = torch.aten.sum.dim_IntList %3377, %3383, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3385 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3386 = torch.aten.view %3384, %3385 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3387 = torch.aten.detach %3386 : !torch.tensor -> !torch.tensor
    %3388 = torch.aten.detach %3387 : !torch.tensor -> !torch.tensor
    %3389 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3390 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3391 = torch.aten._reshape_alias %3379, %3389, %3390 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3392 = torch.aten.t %3382 : !torch.tensor -> !torch.tensor
    %3393 = torch.aten.detach %3392 : !torch.tensor -> !torch.tensor
    %3394 = torch.aten.detach %3393 : !torch.tensor -> !torch.tensor
    %3395 = torch.aten.gelu_backward %3391, %184, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %3396 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %3397 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3398 = torch.aten._reshape_alias %3395, %3396, %3397 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3399 = torch.aten.t %179 : !torch.tensor -> !torch.tensor
    %3400 = torch.aten.mm %3398, %3399 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3401 = torch.aten.t %3398 : !torch.tensor -> !torch.tensor
    %3402 = torch.aten.mm %3401, %181 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3403 = torch.aten.t %3402 : !torch.tensor -> !torch.tensor
    %3404 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3405 = torch.aten.sum.dim_IntList %3398, %3404, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3406 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %3407 = torch.aten.view %3405, %3406 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3408 = torch.aten.detach %3407 : !torch.tensor -> !torch.tensor
    %3409 = torch.aten.detach %3408 : !torch.tensor -> !torch.tensor
    %3410 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3411 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3412 = torch.aten._reshape_alias %3400, %3410, %3411 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3413 = torch.aten.add.Tensor %3365, %3412, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3414 = torch.aten.t %3403 : !torch.tensor -> !torch.tensor
    %3415 = torch.aten.detach %3414 : !torch.tensor -> !torch.tensor
    %3416 = torch.aten.detach %3415 : !torch.tensor -> !torch.tensor
    %3417 = torch.aten.sub.Tensor %177, %result1_10, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3418 = torch.aten.mul.Tensor %3417, %result2_11 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3419 = torch.aten.mul.Tensor %3413, %arg23 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3420 = torch.aten.mul.Scalar %3419, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3421 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3422 = torch.aten.sum.dim_IntList %3419, %3421, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3423 = torch.aten.mul.Tensor %3419, %3418 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3424 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3425 = torch.aten.sum.dim_IntList %3423, %3424, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3426 = torch.aten.mul.Tensor %3418, %3425 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3427 = torch.aten.sub.Tensor %3420, %3422, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3428 = torch.aten.sub.Tensor %3427, %3426, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3429 = torch.aten.div.Scalar %result2_11, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3430 = torch.aten.mul.Tensor %3429, %3428 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3431 = torch.aten.mul.Tensor %3413, %3418 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3432 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3433 = torch.aten.sum.dim_IntList %3431, %3432, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3434 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3435 = torch.aten.sum.dim_IntList %3413, %3434, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3436 = torch.aten.detach %3433 : !torch.tensor -> !torch.tensor
    %3437 = torch.aten.detach %3436 : !torch.tensor -> !torch.tensor
    %3438 = torch.aten.detach %3435 : !torch.tensor -> !torch.tensor
    %3439 = torch.aten.detach %3438 : !torch.tensor -> !torch.tensor
    %3440 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3441 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3442 = torch.aten._reshape_alias %3430, %3440, %3441 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3443 = torch.aten.t %171 : !torch.tensor -> !torch.tensor
    %3444 = torch.aten.mm %3442, %3443 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3445 = torch.aten.t %3442 : !torch.tensor -> !torch.tensor
    %3446 = torch.aten.mm %3445, %173 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3447 = torch.aten.t %3446 : !torch.tensor -> !torch.tensor
    %3448 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3449 = torch.aten.sum.dim_IntList %3442, %3448, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3450 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3451 = torch.aten.view %3449, %3450 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3452 = torch.aten.detach %3451 : !torch.tensor -> !torch.tensor
    %3453 = torch.aten.detach %3452 : !torch.tensor -> !torch.tensor
    %3454 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3455 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3456 = torch.aten._reshape_alias %3444, %3454, %3455 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3457 = torch.aten.t %3447 : !torch.tensor -> !torch.tensor
    %3458 = torch.aten.detach %3457 : !torch.tensor -> !torch.tensor
    %3459 = torch.aten.detach %3458 : !torch.tensor -> !torch.tensor
    %3460 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3461 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3462 = torch.aten._reshape_alias %3456, %3460, %3461 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3463 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3464 = torch.aten.permute %3462, %3463 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3465 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3466 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3467 = torch.aten._reshape_alias %3464, %3465, %3466 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3468 = torch.aten.transpose.int %157, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3469 = torch.aten.bmm %3468, %3467 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3470 = torch.aten.transpose.int %162, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3471 = torch.aten.bmm %3467, %3470 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3472 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3473 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3474 = torch.aten._reshape_alias %3469, %3472, %3473 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3475 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3476 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3477 = torch.aten._reshape_alias %3471, %3475, %3476 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3478 = torch.aten.detach %152 : !torch.tensor -> !torch.tensor
    %3479 = torch.aten._softmax_backward_data %3477, %3478, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3480 = torch.aten.div.Scalar %3479, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %3481 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3482 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3483 = torch.aten._reshape_alias %3480, %3481, %3482 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3484 = torch.aten.transpose.int %139, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3485 = torch.aten.bmm %3484, %3483 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3486 = torch.aten.transpose.int %144, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3487 = torch.aten.bmm %3483, %3486 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3488 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3489 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3490 = torch.aten._reshape_alias %3485, %3488, %3489 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3491 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3492 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3493 = torch.aten._reshape_alias %3487, %3491, %3492 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3494 = torch.aten.transpose.int %3490, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3495 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3496 = torch.aten.permute %3493, %3495 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3497 = torch.aten.clone %3496, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3498 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3499 = torch.aten._unsafe_view %3497, %3498 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3500 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3501 = torch.aten.permute %3474, %3500 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3502 = torch.aten.clone %3501, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3503 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3504 = torch.aten._unsafe_view %3502, %3503 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3505 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3506 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3507 = torch.aten._reshape_alias %3504, %3505, %3506 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3508 = torch.aten.t %120 : !torch.tensor -> !torch.tensor
    %3509 = torch.aten.mm %3507, %3508 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3510 = torch.aten.t %3507 : !torch.tensor -> !torch.tensor
    %3511 = torch.aten.mm %3510, %122 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3512 = torch.aten.t %3511 : !torch.tensor -> !torch.tensor
    %3513 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3514 = torch.aten.sum.dim_IntList %3507, %3513, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3515 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3516 = torch.aten.view %3514, %3515 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3517 = torch.aten.detach %3516 : !torch.tensor -> !torch.tensor
    %3518 = torch.aten.detach %3517 : !torch.tensor -> !torch.tensor
    %3519 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3520 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3521 = torch.aten._reshape_alias %3509, %3519, %3520 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3522 = torch.aten.add.Tensor %3430, %3521, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3523 = torch.aten.t %3512 : !torch.tensor -> !torch.tensor
    %3524 = torch.aten.detach %3523 : !torch.tensor -> !torch.tensor
    %3525 = torch.aten.detach %3524 : !torch.tensor -> !torch.tensor
    %3526 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3527 = torch.aten.permute %3494, %3526 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3528 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3529 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3530 = torch.aten._reshape_alias %3527, %3528, %3529 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3531 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3532 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %3533 = torch.aten._reshape_alias %3530, %3531, %3532 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3534 = torch.aten.t %110 : !torch.tensor -> !torch.tensor
    %3535 = torch.aten.mm %3533, %3534 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3536 = torch.aten.t %3533 : !torch.tensor -> !torch.tensor
    %3537 = torch.aten.mm %3536, %112 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3538 = torch.aten.t %3537 : !torch.tensor -> !torch.tensor
    %3539 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3540 = torch.aten.sum.dim_IntList %3533, %3539, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3541 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3542 = torch.aten.view %3540, %3541 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3543 = torch.aten.detach %3542 : !torch.tensor -> !torch.tensor
    %3544 = torch.aten.detach %3543 : !torch.tensor -> !torch.tensor
    %3545 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3546 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3547 = torch.aten._reshape_alias %3535, %3545, %3546 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3548 = torch.aten.add.Tensor %3522, %3547, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3549 = torch.aten.t %3538 : !torch.tensor -> !torch.tensor
    %3550 = torch.aten.detach %3549 : !torch.tensor -> !torch.tensor
    %3551 = torch.aten.detach %3550 : !torch.tensor -> !torch.tensor
    %3552 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3553 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3554 = torch.aten._reshape_alias %3499, %3552, %3553 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3555 = torch.aten.t %104 : !torch.tensor -> !torch.tensor
    %3556 = torch.aten.mm %3554, %3555 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3557 = torch.aten.t %3554 : !torch.tensor -> !torch.tensor
    %3558 = torch.aten.mm %3557, %106 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3559 = torch.aten.t %3558 : !torch.tensor -> !torch.tensor
    %3560 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3561 = torch.aten.sum.dim_IntList %3554, %3560, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3562 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3563 = torch.aten.view %3561, %3562 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3564 = torch.aten.detach %3563 : !torch.tensor -> !torch.tensor
    %3565 = torch.aten.detach %3564 : !torch.tensor -> !torch.tensor
    %3566 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3567 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3568 = torch.aten._reshape_alias %3556, %3566, %3567 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3569 = torch.aten.add.Tensor %3548, %3568, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3570 = torch.aten.t %3559 : !torch.tensor -> !torch.tensor
    %3571 = torch.aten.detach %3570 : !torch.tensor -> !torch.tensor
    %3572 = torch.aten.detach %3571 : !torch.tensor -> !torch.tensor
    %3573 = torch.aten.sub.Tensor %102, %result1_7, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3574 = torch.aten.mul.Tensor %3573, %result2_8 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3575 = torch.aten.mul.Tensor %3569, %arg19 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3576 = torch.aten.mul.Scalar %3575, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3577 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3578 = torch.aten.sum.dim_IntList %3575, %3577, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3579 = torch.aten.mul.Tensor %3575, %3574 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3580 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3581 = torch.aten.sum.dim_IntList %3579, %3580, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3582 = torch.aten.mul.Tensor %3574, %3581 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3583 = torch.aten.sub.Tensor %3576, %3578, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3584 = torch.aten.sub.Tensor %3583, %3582, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3585 = torch.aten.div.Scalar %result2_8, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3586 = torch.aten.mul.Tensor %3585, %3584 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3587 = torch.aten.mul.Tensor %3569, %3574 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3588 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3589 = torch.aten.sum.dim_IntList %3587, %3588, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3590 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3591 = torch.aten.sum.dim_IntList %3569, %3590, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3592 = torch.aten.detach %3589 : !torch.tensor -> !torch.tensor
    %3593 = torch.aten.detach %3592 : !torch.tensor -> !torch.tensor
    %3594 = torch.aten.detach %3591 : !torch.tensor -> !torch.tensor
    %3595 = torch.aten.detach %3594 : !torch.tensor -> !torch.tensor
    %3596 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3597 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3598 = torch.aten._reshape_alias %3586, %3596, %3597 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3599 = torch.aten.t %96 : !torch.tensor -> !torch.tensor
    %3600 = torch.aten.mm %3598, %3599 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3601 = torch.aten.t %3598 : !torch.tensor -> !torch.tensor
    %3602 = torch.aten.mm %3601, %98 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3603 = torch.aten.t %3602 : !torch.tensor -> !torch.tensor
    %3604 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3605 = torch.aten.sum.dim_IntList %3598, %3604, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3606 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3607 = torch.aten.view %3605, %3606 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3608 = torch.aten.detach %3607 : !torch.tensor -> !torch.tensor
    %3609 = torch.aten.detach %3608 : !torch.tensor -> !torch.tensor
    %3610 = torch.prim.ListConstruct %int1, %int128, %int1536 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3611 = torch.prim.ListConstruct %int196608, %int1536, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3612 = torch.aten._reshape_alias %3600, %3610, %3611 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3613 = torch.aten.t %3603 : !torch.tensor -> !torch.tensor
    %3614 = torch.aten.detach %3613 : !torch.tensor -> !torch.tensor
    %3615 = torch.aten.detach %3614 : !torch.tensor -> !torch.tensor
    %3616 = torch.aten.gelu_backward %3612, %94, %str_2 : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %3617 = torch.prim.ListConstruct %int128, %int1536 : (!torch.int, !torch.int) -> !torch.list<int>
    %3618 = torch.prim.ListConstruct %int1536, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3619 = torch.aten._reshape_alias %3616, %3617, %3618 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3620 = torch.aten.t %89 : !torch.tensor -> !torch.tensor
    %3621 = torch.aten.mm %3619, %3620 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3622 = torch.aten.t %3619 : !torch.tensor -> !torch.tensor
    %3623 = torch.aten.mm %3622, %91 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3624 = torch.aten.t %3623 : !torch.tensor -> !torch.tensor
    %3625 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3626 = torch.aten.sum.dim_IntList %3619, %3625, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3627 = torch.prim.ListConstruct %int1536 : (!torch.int) -> !torch.list<int>
    %3628 = torch.aten.view %3626, %3627 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3629 = torch.aten.detach %3628 : !torch.tensor -> !torch.tensor
    %3630 = torch.aten.detach %3629 : !torch.tensor -> !torch.tensor
    %3631 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3632 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3633 = torch.aten._reshape_alias %3621, %3631, %3632 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3634 = torch.aten.add.Tensor %3586, %3633, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3635 = torch.aten.t %3624 : !torch.tensor -> !torch.tensor
    %3636 = torch.aten.detach %3635 : !torch.tensor -> !torch.tensor
    %3637 = torch.aten.detach %3636 : !torch.tensor -> !torch.tensor
    %3638 = torch.aten.sub.Tensor %87, %result1_4, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3639 = torch.aten.mul.Tensor %3638, %result2_5 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3640 = torch.aten.mul.Tensor %3634, %arg7 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3641 = torch.aten.mul.Scalar %3640, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3642 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3643 = torch.aten.sum.dim_IntList %3640, %3642, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3644 = torch.aten.mul.Tensor %3640, %3639 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3645 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3646 = torch.aten.sum.dim_IntList %3644, %3645, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3647 = torch.aten.mul.Tensor %3639, %3646 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3648 = torch.aten.sub.Tensor %3641, %3643, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3649 = torch.aten.sub.Tensor %3648, %3647, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3650 = torch.aten.div.Scalar %result2_5, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3651 = torch.aten.mul.Tensor %3650, %3649 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3652 = torch.aten.mul.Tensor %3634, %3639 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3653 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3654 = torch.aten.sum.dim_IntList %3652, %3653, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3655 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3656 = torch.aten.sum.dim_IntList %3634, %3655, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3657 = torch.aten.detach %3654 : !torch.tensor -> !torch.tensor
    %3658 = torch.aten.detach %3657 : !torch.tensor -> !torch.tensor
    %3659 = torch.aten.detach %3656 : !torch.tensor -> !torch.tensor
    %3660 = torch.aten.detach %3659 : !torch.tensor -> !torch.tensor
    %3661 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3662 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3663 = torch.aten._reshape_alias %3651, %3661, %3662 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3664 = torch.aten.t %81 : !torch.tensor -> !torch.tensor
    %3665 = torch.aten.mm %3663, %3664 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3666 = torch.aten.t %3663 : !torch.tensor -> !torch.tensor
    %3667 = torch.aten.mm %3666, %83 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3668 = torch.aten.t %3667 : !torch.tensor -> !torch.tensor
    %3669 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3670 = torch.aten.sum.dim_IntList %3663, %3669, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3671 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3672 = torch.aten.view %3670, %3671 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3673 = torch.aten.detach %3672 : !torch.tensor -> !torch.tensor
    %3674 = torch.aten.detach %3673 : !torch.tensor -> !torch.tensor
    %3675 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3676 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3677 = torch.aten._reshape_alias %3665, %3675, %3676 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3678 = torch.aten.t %3668 : !torch.tensor -> !torch.tensor
    %3679 = torch.aten.detach %3678 : !torch.tensor -> !torch.tensor
    %3680 = torch.aten.detach %3679 : !torch.tensor -> !torch.tensor
    %3681 = torch.prim.ListConstruct %int1, %int128, %int12, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3682 = torch.prim.ListConstruct %int49152, %int384, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3683 = torch.aten._reshape_alias %3677, %3681, %3682 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3684 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3685 = torch.aten.permute %3683, %3684 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3686 = torch.prim.ListConstruct %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3687 = torch.prim.ListConstruct %int32, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3688 = torch.aten._reshape_alias %3685, %3686, %3687 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3689 = torch.aten.transpose.int %67, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3690 = torch.aten.bmm %3689, %3688 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3691 = torch.aten.transpose.int %72, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3692 = torch.aten.bmm %3688, %3691 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3693 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3694 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3695 = torch.aten._reshape_alias %3690, %3693, %3694 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3696 = torch.prim.ListConstruct %int1, %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3697 = torch.prim.ListConstruct %int196608, %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3698 = torch.aten._reshape_alias %3692, %3696, %3697 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3699 = torch.aten.detach %62 : !torch.tensor -> !torch.tensor
    %3700 = torch.aten._softmax_backward_data %3698, %3699, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3701 = torch.aten.div.Scalar %3700, %float5.656850e00 : !torch.tensor, !torch.float -> !torch.tensor
    %3702 = torch.prim.ListConstruct %int12, %int128, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3703 = torch.prim.ListConstruct %int16384, %int128, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3704 = torch.aten._reshape_alias %3701, %3702, %3703 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3705 = torch.aten.transpose.int %49, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3706 = torch.aten.bmm %3705, %3704 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3707 = torch.aten.transpose.int %54, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3708 = torch.aten.bmm %3704, %3707 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3709 = torch.prim.ListConstruct %int1, %int12, %int32, %int128 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3710 = torch.prim.ListConstruct %int49152, %int4096, %int128, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3711 = torch.aten._reshape_alias %3706, %3709, %3710 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3712 = torch.prim.ListConstruct %int1, %int12, %int128, %int32 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3713 = torch.prim.ListConstruct %int49152, %int4096, %int32, %int1 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3714 = torch.aten._reshape_alias %3708, %3712, %3713 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3715 = torch.aten.transpose.int %3711, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %3716 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3717 = torch.aten.permute %3714, %3716 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3718 = torch.aten.clone %3717, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3719 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3720 = torch.aten._unsafe_view %3718, %3719 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3721 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3722 = torch.aten.permute %3695, %3721 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3723 = torch.aten.clone %3722, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3724 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3725 = torch.aten._unsafe_view %3723, %3724 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3726 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3727 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3728 = torch.aten._reshape_alias %3725, %3726, %3727 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3729 = torch.aten.t %30 : !torch.tensor -> !torch.tensor
    %3730 = torch.aten.mm %3728, %3729 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3731 = torch.aten.t %3728 : !torch.tensor -> !torch.tensor
    %3732 = torch.aten.mm %3731, %32 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3733 = torch.aten.t %3732 : !torch.tensor -> !torch.tensor
    %3734 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3735 = torch.aten.sum.dim_IntList %3728, %3734, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3736 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3737 = torch.aten.view %3735, %3736 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3738 = torch.aten.detach %3737 : !torch.tensor -> !torch.tensor
    %3739 = torch.aten.detach %3738 : !torch.tensor -> !torch.tensor
    %3740 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3741 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3742 = torch.aten._reshape_alias %3730, %3740, %3741 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3743 = torch.aten.add.Tensor %3651, %3742, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3744 = torch.aten.t %3733 : !torch.tensor -> !torch.tensor
    %3745 = torch.aten.detach %3744 : !torch.tensor -> !torch.tensor
    %3746 = torch.aten.detach %3745 : !torch.tensor -> !torch.tensor
    %3747 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3748 = torch.aten.permute %3715, %3747 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3749 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3750 = torch.prim.ListConstruct %int128, %int1, %int128 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3751 = torch.aten._reshape_alias %3748, %3749, %3750 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3752 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3753 = torch.prim.ListConstruct %int1, %int128 : (!torch.int, !torch.int) -> !torch.list<int>
    %3754 = torch.aten._reshape_alias %3751, %3752, %3753 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3755 = torch.aten.t %20 : !torch.tensor -> !torch.tensor
    %3756 = torch.aten.mm %3754, %3755 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3757 = torch.aten.t %3754 : !torch.tensor -> !torch.tensor
    %3758 = torch.aten.mm %3757, %22 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3759 = torch.aten.t %3758 : !torch.tensor -> !torch.tensor
    %3760 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3761 = torch.aten.sum.dim_IntList %3754, %3760, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3762 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3763 = torch.aten.view %3761, %3762 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3764 = torch.aten.detach %3763 : !torch.tensor -> !torch.tensor
    %3765 = torch.aten.detach %3764 : !torch.tensor -> !torch.tensor
    %3766 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3767 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3768 = torch.aten._reshape_alias %3756, %3766, %3767 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3769 = torch.aten.add.Tensor %3743, %3768, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3770 = torch.aten.t %3759 : !torch.tensor -> !torch.tensor
    %3771 = torch.aten.detach %3770 : !torch.tensor -> !torch.tensor
    %3772 = torch.aten.detach %3771 : !torch.tensor -> !torch.tensor
    %3773 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3774 = torch.prim.ListConstruct %int384, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3775 = torch.aten._reshape_alias %3720, %3773, %3774 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3776 = torch.aten.t %14 : !torch.tensor -> !torch.tensor
    %3777 = torch.aten.mm %3775, %3776 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3778 = torch.aten.t %3775 : !torch.tensor -> !torch.tensor
    %3779 = torch.aten.mm %3778, %16 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3780 = torch.aten.t %3779 : !torch.tensor -> !torch.tensor
    %3781 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %3782 = torch.aten.sum.dim_IntList %3775, %3781, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3783 = torch.prim.ListConstruct %int384 : (!torch.int) -> !torch.list<int>
    %3784 = torch.aten.view %3782, %3783 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3785 = torch.aten.detach %3784 : !torch.tensor -> !torch.tensor
    %3786 = torch.aten.detach %3785 : !torch.tensor -> !torch.tensor
    %3787 = torch.prim.ListConstruct %int1, %int128, %int384 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3788 = torch.prim.ListConstruct %int49152, %int384, %int1 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %3789 = torch.aten._reshape_alias %3777, %3787, %3788 : !torch.tensor, !torch.list<int>, !torch.list<int> -> !torch.tensor
    %3790 = torch.aten.add.Tensor %3769, %3789, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3791 = torch.aten.t %3780 : !torch.tensor -> !torch.tensor
    %3792 = torch.aten.detach %3791 : !torch.tensor -> !torch.tensor
    %3793 = torch.aten.detach %3792 : !torch.tensor -> !torch.tensor
    %3794 = torch.aten.sub.Tensor %12, %result1, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3795 = torch.aten.mul.Tensor %3794, %result2 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3796 = torch.aten.mul.Tensor %3790, %arg2 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3797 = torch.aten.mul.Scalar %3796, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3798 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3799 = torch.aten.sum.dim_IntList %3796, %3798, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3800 = torch.aten.mul.Tensor %3796, %3795 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3801 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %3802 = torch.aten.sum.dim_IntList %3800, %3801, %true_0, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3803 = torch.aten.mul.Tensor %3795, %3802 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3804 = torch.aten.sub.Tensor %3797, %3799, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3805 = torch.aten.sub.Tensor %3804, %3803, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %3806 = torch.aten.div.Scalar %result2, %int384 : !torch.tensor, !torch.int -> !torch.tensor
    %3807 = torch.aten.mul.Tensor %3806, %3805 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3808 = torch.aten.mul.Tensor %3790, %3795 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3809 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3810 = torch.aten.sum.dim_IntList %3808, %3809, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3811 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %3812 = torch.aten.sum.dim_IntList %3790, %3811, %false, %none_1 : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %3813 = torch.aten.detach %3810 : !torch.tensor -> !torch.tensor
    %3814 = torch.aten.detach %3813 : !torch.tensor -> !torch.tensor
    %3815 = torch.aten.detach %3812 : !torch.tensor -> !torch.tensor
    %3816 = torch.aten.detach %3815 : !torch.tensor -> !torch.tensor
    %3817 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3818 = torch.aten.view %3807, %3817 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3819 = torch.prim.ListConstruct %int512, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3820 = torch.aten.new_zeros %3807, %3819, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %3821 = torch.prim.ListConstruct %int128 : (!torch.int) -> !torch.list<int>
    %3822 = torch.aten.view %7, %3821 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3823 = torch.aten.ne.Scalar %3822, %int-1 : !torch.tensor, !torch.int -> !torch.tensor
    %3824 = torch.aten.unsqueeze %3823, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %3825 = torch.aten.expand_as %3824, %3818 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3826 = torch.aten.full_like %3818, %int0, %int6, %int0, %cpu, %false, %none_1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.none -> !torch.tensor
    %3827 = torch.aten.where.self %3825, %3818, %3826 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %3828 = torch.prim.ListConstruct %3822 : (!torch.tensor) -> !torch.list<tensor>
    %3829 = torch.aten.index_put.hacked_twin %3820, %3828, %3827, %true_0 : !torch.tensor, !torch.list<tensor>, !torch.tensor, !torch.bool -> !torch.tensor
    %3830 = torch.aten.detach %3829 : !torch.tensor -> !torch.tensor
    %3831 = torch.aten.detach %3830 : !torch.tensor -> !torch.tensor
    %3832 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3833 = torch.aten.view %3807, %3832 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3834 = torch.prim.ListConstruct %int2, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3835 = torch.aten.new_zeros %3807, %3834, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %3836 = torch.prim.ListConstruct %int128 : (!torch.int) -> !torch.list<int>
    %3837 = torch.aten.view %5, %3836 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3838 = torch.aten.ne.Scalar %3837, %int-1 : !torch.tensor, !torch.int -> !torch.tensor
    %3839 = torch.aten.unsqueeze %3838, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %3840 = torch.aten.expand_as %3839, %3833 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3841 = torch.aten.full_like %3833, %int0, %int6, %int0, %cpu, %false, %none_1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.none -> !torch.tensor
    %3842 = torch.aten.where.self %3840, %3833, %3841 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %3843 = torch.prim.ListConstruct %3837 : (!torch.tensor) -> !torch.list<tensor>
    %3844 = torch.aten.index_put.hacked_twin %3835, %3843, %3842, %true_0 : !torch.tensor, !torch.list<tensor>, !torch.tensor, !torch.bool -> !torch.tensor
    %3845 = torch.aten.detach %3844 : !torch.tensor -> !torch.tensor
    %3846 = torch.aten.detach %3845 : !torch.tensor -> !torch.tensor
    %3847 = torch.prim.ListConstruct %int128, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3848 = torch.aten.view %3807, %3847 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3849 = torch.prim.ListConstruct %int30522, %int384 : (!torch.int, !torch.int) -> !torch.list<int>
    %3850 = torch.aten.new_zeros %3807, %3849, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %3851 = torch.prim.ListConstruct %int128 : (!torch.int) -> !torch.list<int>
    %3852 = torch.aten.view %arg204, %3851 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %3853 = torch.aten.ne.Scalar %3852, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %3854 = torch.aten.unsqueeze %3853, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %3855 = torch.aten.expand_as %3854, %3848 : !torch.tensor, !torch.tensor -> !torch.tensor
    %3856 = torch.aten.full_like %3848, %int0, %int6, %int0, %cpu, %false, %none_1 : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.none -> !torch.tensor
    %3857 = torch.aten.where.self %3855, %3848, %3856 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %3858 = torch.prim.ListConstruct %3852 : (!torch.tensor) -> !torch.list<tensor>
    %3859 = torch.aten.index_put.hacked_twin %3850, %3858, %3857, %true_0 : !torch.tensor, !torch.list<tensor>, !torch.tensor, !torch.bool -> !torch.tensor
    %3860 = torch.aten.detach %3859 : !torch.tensor -> !torch.tensor
    %3861 = torch.aten.detach %3860 : !torch.tensor -> !torch.tensor
    %3862 = torch.aten.add_.Tensor %arg1, %3816, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3863 = torch.aten.add_.Tensor %arg2, %3814, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3864 = torch.aten.add_.Tensor %arg3, %3831, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3865 = torch.aten.add_.Tensor %arg4, %3846, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3866 = torch.aten.add_.Tensor %arg5, %3861, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3867 = torch.aten.add_.Tensor %arg6, %3660, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3868 = torch.aten.add_.Tensor %arg7, %3658, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3869 = torch.aten.add_.Tensor %arg8, %3674, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3870 = torch.aten.add_.Tensor %arg9, %3680, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3871 = torch.aten.add_.Tensor %arg10, %3765, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3872 = torch.aten.add_.Tensor %arg11, %3772, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3873 = torch.aten.add_.Tensor %arg12, %3786, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3874 = torch.aten.add_.Tensor %arg13, %3793, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3875 = torch.aten.add_.Tensor %arg14, %3739, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3876 = torch.aten.add_.Tensor %arg15, %3746, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3877 = torch.aten.add_.Tensor %arg16, %3630, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3878 = torch.aten.add_.Tensor %arg17, %3637, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3879 = torch.aten.add_.Tensor %arg18, %3595, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3880 = torch.aten.add_.Tensor %arg19, %3593, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3881 = torch.aten.add_.Tensor %arg20, %3609, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3882 = torch.aten.add_.Tensor %arg21, %3615, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3883 = torch.aten.add_.Tensor %arg22, %3439, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3884 = torch.aten.add_.Tensor %arg23, %3437, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3885 = torch.aten.add_.Tensor %arg24, %3453, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3886 = torch.aten.add_.Tensor %arg25, %3459, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3887 = torch.aten.add_.Tensor %arg26, %3544, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3888 = torch.aten.add_.Tensor %arg27, %3551, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3889 = torch.aten.add_.Tensor %arg28, %3565, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3890 = torch.aten.add_.Tensor %arg29, %3572, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3891 = torch.aten.add_.Tensor %arg30, %3518, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3892 = torch.aten.add_.Tensor %arg31, %3525, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3893 = torch.aten.add_.Tensor %arg32, %3409, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3894 = torch.aten.add_.Tensor %arg33, %3416, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3895 = torch.aten.add_.Tensor %arg34, %3374, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3896 = torch.aten.add_.Tensor %arg35, %3372, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3897 = torch.aten.add_.Tensor %arg36, %3388, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3898 = torch.aten.add_.Tensor %arg37, %3394, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3899 = torch.aten.add_.Tensor %arg38, %1450, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3900 = torch.aten.add_.Tensor %arg39, %1448, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3901 = torch.aten.add_.Tensor %arg40, %1464, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3902 = torch.aten.add_.Tensor %arg41, %1470, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3903 = torch.aten.add_.Tensor %arg42, %1555, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3904 = torch.aten.add_.Tensor %arg43, %1562, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3905 = torch.aten.add_.Tensor %arg44, %1576, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3906 = torch.aten.add_.Tensor %arg45, %1583, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3907 = torch.aten.add_.Tensor %arg46, %1529, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3908 = torch.aten.add_.Tensor %arg47, %1536, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3909 = torch.aten.add_.Tensor %arg48, %1420, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3910 = torch.aten.add_.Tensor %arg49, %1427, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3911 = torch.aten.add_.Tensor %arg50, %1385, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3912 = torch.aten.add_.Tensor %arg51, %1383, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3913 = torch.aten.add_.Tensor %arg52, %1399, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3914 = torch.aten.add_.Tensor %arg53, %1405, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3915 = torch.aten.add_.Tensor %arg54, %1229, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3916 = torch.aten.add_.Tensor %arg55, %1227, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3917 = torch.aten.add_.Tensor %arg56, %1243, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3918 = torch.aten.add_.Tensor %arg57, %1249, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3919 = torch.aten.add_.Tensor %arg58, %1334, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3920 = torch.aten.add_.Tensor %arg59, %1341, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3921 = torch.aten.add_.Tensor %arg60, %1355, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3922 = torch.aten.add_.Tensor %arg61, %1362, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3923 = torch.aten.add_.Tensor %arg62, %1308, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3924 = torch.aten.add_.Tensor %arg63, %1315, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3925 = torch.aten.add_.Tensor %arg64, %1199, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3926 = torch.aten.add_.Tensor %arg65, %1206, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3927 = torch.aten.add_.Tensor %arg66, %1164, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3928 = torch.aten.add_.Tensor %arg67, %1162, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3929 = torch.aten.add_.Tensor %arg68, %1178, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3930 = torch.aten.add_.Tensor %arg69, %1184, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3931 = torch.aten.add_.Tensor %arg70, %3218, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3932 = torch.aten.add_.Tensor %arg71, %3216, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3933 = torch.aten.add_.Tensor %arg72, %3232, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3934 = torch.aten.add_.Tensor %arg73, %3238, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3935 = torch.aten.add_.Tensor %arg74, %3323, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3936 = torch.aten.add_.Tensor %arg75, %3330, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3937 = torch.aten.add_.Tensor %arg76, %3344, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3938 = torch.aten.add_.Tensor %arg77, %3351, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3939 = torch.aten.add_.Tensor %arg78, %3297, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3940 = torch.aten.add_.Tensor %arg79, %3304, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3941 = torch.aten.add_.Tensor %arg80, %3188, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3942 = torch.aten.add_.Tensor %arg81, %3195, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3943 = torch.aten.add_.Tensor %arg82, %3153, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3944 = torch.aten.add_.Tensor %arg83, %3151, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3945 = torch.aten.add_.Tensor %arg84, %3167, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3946 = torch.aten.add_.Tensor %arg85, %3173, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3947 = torch.aten.add_.Tensor %arg86, %2997, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3948 = torch.aten.add_.Tensor %arg87, %2995, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3949 = torch.aten.add_.Tensor %arg88, %3011, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3950 = torch.aten.add_.Tensor %arg89, %3017, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3951 = torch.aten.add_.Tensor %arg90, %3102, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3952 = torch.aten.add_.Tensor %arg91, %3109, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3953 = torch.aten.add_.Tensor %arg92, %3123, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3954 = torch.aten.add_.Tensor %arg93, %3130, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3955 = torch.aten.add_.Tensor %arg94, %3076, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3956 = torch.aten.add_.Tensor %arg95, %3083, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3957 = torch.aten.add_.Tensor %arg96, %2967, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3958 = torch.aten.add_.Tensor %arg97, %2974, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3959 = torch.aten.add_.Tensor %arg98, %2932, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3960 = torch.aten.add_.Tensor %arg99, %2930, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3961 = torch.aten.add_.Tensor %arg100, %2946, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3962 = torch.aten.add_.Tensor %arg101, %2952, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3963 = torch.aten.add_.Tensor %arg102, %2776, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3964 = torch.aten.add_.Tensor %arg103, %2774, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3965 = torch.aten.add_.Tensor %arg104, %2790, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3966 = torch.aten.add_.Tensor %arg105, %2796, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3967 = torch.aten.add_.Tensor %arg106, %2881, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3968 = torch.aten.add_.Tensor %arg107, %2888, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3969 = torch.aten.add_.Tensor %arg108, %2902, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3970 = torch.aten.add_.Tensor %arg109, %2909, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3971 = torch.aten.add_.Tensor %arg110, %2855, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3972 = torch.aten.add_.Tensor %arg111, %2862, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3973 = torch.aten.add_.Tensor %arg112, %2746, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3974 = torch.aten.add_.Tensor %arg113, %2753, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3975 = torch.aten.add_.Tensor %arg114, %2711, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3976 = torch.aten.add_.Tensor %arg115, %2709, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3977 = torch.aten.add_.Tensor %arg116, %2725, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3978 = torch.aten.add_.Tensor %arg117, %2731, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3979 = torch.aten.add_.Tensor %arg118, %2555, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3980 = torch.aten.add_.Tensor %arg119, %2553, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3981 = torch.aten.add_.Tensor %arg120, %2569, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3982 = torch.aten.add_.Tensor %arg121, %2575, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3983 = torch.aten.add_.Tensor %arg122, %2660, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3984 = torch.aten.add_.Tensor %arg123, %2667, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3985 = torch.aten.add_.Tensor %arg124, %2681, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3986 = torch.aten.add_.Tensor %arg125, %2688, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3987 = torch.aten.add_.Tensor %arg126, %2634, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3988 = torch.aten.add_.Tensor %arg127, %2641, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3989 = torch.aten.add_.Tensor %arg128, %2525, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3990 = torch.aten.add_.Tensor %arg129, %2532, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3991 = torch.aten.add_.Tensor %arg130, %2490, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3992 = torch.aten.add_.Tensor %arg131, %2488, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3993 = torch.aten.add_.Tensor %arg132, %2504, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3994 = torch.aten.add_.Tensor %arg133, %2510, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3995 = torch.aten.add_.Tensor %arg134, %2334, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3996 = torch.aten.add_.Tensor %arg135, %2332, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3997 = torch.aten.add_.Tensor %arg136, %2348, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3998 = torch.aten.add_.Tensor %arg137, %2354, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %3999 = torch.aten.add_.Tensor %arg138, %2439, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4000 = torch.aten.add_.Tensor %arg139, %2446, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4001 = torch.aten.add_.Tensor %arg140, %2460, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4002 = torch.aten.add_.Tensor %arg141, %2467, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4003 = torch.aten.add_.Tensor %arg142, %2413, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4004 = torch.aten.add_.Tensor %arg143, %2420, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4005 = torch.aten.add_.Tensor %arg144, %2304, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4006 = torch.aten.add_.Tensor %arg145, %2311, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4007 = torch.aten.add_.Tensor %arg146, %2269, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4008 = torch.aten.add_.Tensor %arg147, %2267, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4009 = torch.aten.add_.Tensor %arg148, %2283, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4010 = torch.aten.add_.Tensor %arg149, %2289, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4011 = torch.aten.add_.Tensor %arg150, %2113, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4012 = torch.aten.add_.Tensor %arg151, %2111, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4013 = torch.aten.add_.Tensor %arg152, %2127, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4014 = torch.aten.add_.Tensor %arg153, %2133, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4015 = torch.aten.add_.Tensor %arg154, %2218, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4016 = torch.aten.add_.Tensor %arg155, %2225, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4017 = torch.aten.add_.Tensor %arg156, %2239, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4018 = torch.aten.add_.Tensor %arg157, %2246, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4019 = torch.aten.add_.Tensor %arg158, %2192, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4020 = torch.aten.add_.Tensor %arg159, %2199, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4021 = torch.aten.add_.Tensor %arg160, %2083, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4022 = torch.aten.add_.Tensor %arg161, %2090, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4023 = torch.aten.add_.Tensor %arg162, %2048, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4024 = torch.aten.add_.Tensor %arg163, %2046, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4025 = torch.aten.add_.Tensor %arg164, %2062, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4026 = torch.aten.add_.Tensor %arg165, %2068, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4027 = torch.aten.add_.Tensor %arg166, %1892, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4028 = torch.aten.add_.Tensor %arg167, %1890, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4029 = torch.aten.add_.Tensor %arg168, %1906, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4030 = torch.aten.add_.Tensor %arg169, %1912, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4031 = torch.aten.add_.Tensor %arg170, %1997, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4032 = torch.aten.add_.Tensor %arg171, %2004, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4033 = torch.aten.add_.Tensor %arg172, %2018, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4034 = torch.aten.add_.Tensor %arg173, %2025, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4035 = torch.aten.add_.Tensor %arg174, %1971, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4036 = torch.aten.add_.Tensor %arg175, %1978, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4037 = torch.aten.add_.Tensor %arg176, %1862, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4038 = torch.aten.add_.Tensor %arg177, %1869, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4039 = torch.aten.add_.Tensor %arg178, %1827, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4040 = torch.aten.add_.Tensor %arg179, %1825, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4041 = torch.aten.add_.Tensor %arg180, %1841, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4042 = torch.aten.add_.Tensor %arg181, %1847, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4043 = torch.aten.add_.Tensor %arg182, %1671, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4044 = torch.aten.add_.Tensor %arg183, %1669, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4045 = torch.aten.add_.Tensor %arg184, %1685, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4046 = torch.aten.add_.Tensor %arg185, %1691, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4047 = torch.aten.add_.Tensor %arg186, %1776, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4048 = torch.aten.add_.Tensor %arg187, %1783, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4049 = torch.aten.add_.Tensor %arg188, %1797, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4050 = torch.aten.add_.Tensor %arg189, %1804, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4051 = torch.aten.add_.Tensor %arg190, %1750, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4052 = torch.aten.add_.Tensor %arg191, %1757, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4053 = torch.aten.add_.Tensor %arg192, %1641, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4054 = torch.aten.add_.Tensor %arg193, %1648, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4055 = torch.aten.add_.Tensor %arg194, %1606, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4056 = torch.aten.add_.Tensor %arg195, %1604, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4057 = torch.aten.add_.Tensor %arg196, %1620, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4058 = torch.aten.add_.Tensor %arg197, %1626, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4059 = torch.aten.add_.Tensor %arg198, %1132, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4060 = torch.aten.add_.Tensor %arg199, %1135, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4061 = torch.aten.add_.Tensor %arg200, %1116, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4062 = torch.aten.add_.Tensor %arg201, %1119, %float-1.000000e-02 : !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor
    %4063 = torch.prim.TupleConstruct %3862, %3863, %3864, %3865, %3866, %3867, %3868, %3869, %3870, %3871, %3872, %3873, %3874, %3875, %3876, %3877, %3878, %3879, %3880, %3881, %3882, %3883, %3884, %3885, %3886, %3887, %3888, %3889, %3890, %3891, %3892, %3893, %3894, %3895, %3896, %3897, %3898, %3899, %3900, %3901, %3902, %3903, %3904, %3905, %3906, %3907, %3908, %3909, %3910, %3911, %3912, %3913, %3914, %3915, %3916, %3917, %3918, %3919, %3920, %3921, %3922, %3923, %3924, %3925, %3926, %3927, %3928, %3929, %3930, %3931, %3932, %3933, %3934, %3935, %3936, %3937, %3938, %3939, %3940, %3941, %3942, %3943, %3944, %3945, %3946, %3947, %3948, %3949, %3950, %3951, %3952, %3953, %3954, %3955, %3956, %3957, %3958, %3959, %3960, %3961, %3962, %3963, %3964, %3965, %3966, %3967, %3968, %3969, %3970, %3971, %3972, %3973, %3974, %3975, %3976, %3977, %3978, %3979, %3980, %3981, %3982, %3983, %3984, %3985, %3986, %3987, %3988, %3989, %3990, %3991, %3992, %3993, %3994, %3995, %3996, %3997, %3998, %3999, %4000, %4001, %4002, %4003, %4004, %4005, %4006, %4007, %4008, %4009, %4010, %4011, %4012, %4013, %4014, %4015, %4016, %4017, %4018, %4019, %4020, %4021, %4022, %4023, %4024, %4025, %4026, %4027, %4028, %4029, %4030, %4031, %4032, %4033, %4034, %4035, %4036, %4037, %4038, %4039, %4040, %4041, %4042, %4043, %4044, %4045, %4046, %4047, %4048, %4049, %4050, %4051, %4052, %4053, %4054, %4055, %4056, %4057, %4058, %4059, %4060, %4061, %4062, %arg202, %arg203 : !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tuple<tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor>
    return %4063 : !torch.tuple<tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor, tensor>
  }
  torch.class_type @__torch__.torch.fx.graph_module.forward {
    torch.attr private "_tensor_constant0" : !torch.tensor
    torch.attr private "training" : !torch.bool
    torch.attr private "_is_full_backward_hook" : !torch.optional<bool>
    torch.attr private "_code" : !torch.str
    torch.method private "__code_getter", @__torch__.torch.fx.graph_module.forward.__code_getter
    torch.method "forward", @__torch__.torch.fx.graph_module.forward.forward
  }
  %0 = torch.tensor.literal(dense<-0.000000e+00> : tensor<1x1x1x128xf32>) : !torch.tensor<[1,1,1,128],f32>
  %true = torch.constant.bool true
  %none = torch.constant.none
  %str = torch.constant.str "\0A\0A\0Adef forward(self, params_1, params_2, params_3, params_4, params_5, params_6, params_7, params_8, params_9, params_10, params_11, params_12, params_13, params_14, params_15, params_16, params_17, params_18, params_19, params_20, params_21, params_22, params_23, params_24, params_25, params_26, params_27, params_28, params_29, params_30, params_31, params_32, params_33, params_34, params_35, params_36, params_37, params_38, params_39, params_40, params_41, params_42, params_43, params_44, params_45, params_46, params_47, params_48, params_49, params_50, params_51, params_52, params_53, params_54, params_55, params_56, params_57, params_58, params_59, params_60, params_61, params_62, params_63, params_64, params_65, params_66, params_67, params_68, params_69, params_70, params_71, params_72, params_73, params_74, params_75, params_76, params_77, params_78, params_79, params_80, params_81, params_82, params_83, params_84, params_85, params_86, params_87, params_88, params_89, params_90, params_91, params_92, params_93, params_94, params_95, params_96, params_97, params_98, params_99, params_100, params_101, params_102, params_103, params_104, params_105, params_106, params_107, params_108, params_109, params_110, params_111, params_112, params_113, params_114, params_115, params_116, params_117, params_118, params_119, params_120, params_121, params_122, params_123, params_124, params_125, params_126, params_127, params_128, params_129, params_130, params_131, params_132, params_133, params_134, params_135, params_136, params_137, params_138, params_139, params_140, params_141, params_142, params_143, params_144, params_145, params_146, params_147, params_148, params_149, params_150, params_151, params_152, params_153, params_154, params_155, params_156, params_157, params_158, params_159, params_160, params_161, params_162, params_163, params_164, params_165, params_166, params_167, params_168, params_169, params_170, params_171, params_172, params_173, params_174, params_175, params_176, params_177, params_178, params_179, params_180, params_181, params_182, params_183, params_184, params_185, params_186, params_187, params_188, params_189, params_190, params_191, params_192, params_193, params_194, params_195, params_196, params_197, params_198, params_199, params_200, params_201, buffers_1, buffers_2, args_1):\0A    slice_1 = torch.ops.aten.slice(buffers_2, 0, 0, 9223372036854775807)\0A    slice_2 = torch.ops.aten.slice(slice_1, 1, 0, 128);  slice_1 = None\0A    expand = torch.ops.aten.expand(slice_2, [1, 128]);  slice_2 = None\0A    slice_3 = torch.ops.aten.slice(buffers_1, 0, 0, 9223372036854775807)\0A    slice_4 = torch.ops.aten.slice(slice_3, 1, 0, 128);  slice_3 = None\0A    embedding = torch.ops.aten.embedding(params_5, args_1, 0)\0A    embedding_1 = torch.ops.aten.embedding(params_4, expand)\0A    add = torch.ops.aten.add(embedding, embedding_1);  embedding = embedding_1 = None\0A    embedding_2 = torch.ops.aten.embedding(params_3, slice_4)\0A    add_ = torch.ops.aten.add_(add, embedding_2);  add = embedding_2 = None\0A    native_layer_norm = torch.ops.aten.native_layer_norm(add_, [384], params_2, params_1, 1e-12)\0A    getitem = native_layer_norm[0]\0A    getitem_1 = native_layer_norm[1]\0A    getitem_2 = native_layer_norm[2];  native_layer_norm = None\0A    t = torch.ops.aten.t(params_13)\0A    view = torch.ops.aten.view(getitem, [128, 384])\0A    addmm = torch.ops.aten.addmm(params_12, view, t)\0A    view_1 = torch.ops.aten.view(addmm, [1, 128, 384]);  addmm = None\0A    t_1 = torch.ops.aten.t(params_11)\0A    view_2 = torch.ops.aten.view(getitem, [128, 384])\0A    addmm_1 = torch.ops.aten.addmm(params_10, view_2, t_1)\0A    view_3 = torch.ops.aten.view(addmm_1, [1, 128, 384]);  addmm_1 = None\0A    view_4 = torch.ops.aten.view(view_3, [1, 128, 12, 32]);  view_3 = None\0A    permute = torch.ops.aten.permute(view_4, [0, 2, 1, 3]);  view_4 = None\0A    t_2 = torch.ops.aten.t(params_15)\0A    view_5 = torch.ops.aten.view(getitem, [128, 384])\0A    addmm_2 = torch.ops.aten.addmm(params_14, view_5, t_2)\0A    view_6 = torch.ops.aten.view(addmm_2, [1, 128, 384]);  addmm_2 = None\0A    view_7 = torch.ops.aten.view(view_6, [1, 128, 12, 32]);  view_6 = None\0A    permute_1 = torch.ops.aten.permute(view_7, [0, 2, 1, 3]);  view_7 = None\0A    view_8 = torch.ops.aten.view(view_1, [1, 128, 12, 32]);  view_1 = None\0A    permute_2 = torch.ops.aten.permute(view_8, [0, 2, 1, 3]);  view_8 = None\0A    transpose = torch.ops.aten.transpose(permute, -1, -2);  permute = None\0A    expand_1 = torch.ops.aten.expand(permute_2, [1, 12, 128, 32]);  permute_2 = None\0A    _reshape_alias = torch.ops.aten._reshape_alias(expand_1, [12, 128, 32], [32, 384, 1]);  expand_1 = None\0A    expand_2 = torch.ops.aten.expand(transpose, [1, 12, 32, 128]);  transpose = None\0A    _reshape_alias_1 = torch.ops.aten._reshape_alias(expand_2, [12, 32, 128], [32, 1, 384]);  expand_2 = None\0A    bmm = torch.ops.aten.bmm(_reshape_alias, _reshape_alias_1)\0A    _unsafe_view = torch.ops.aten._unsafe_view(bmm, [1, 12, 128, 128]);  bmm = None\0A    div = torch.ops.aten.div(_unsafe_view, 5.656854249492381);  _unsafe_view = None\0A    _tensor_constant0 = self._tensor_constant0\0A    add_1 = torch.ops.aten.add(div, _tensor_constant0);  div = _tensor_constant0 = None\0A    _softmax = torch.ops.aten._softmax(add_1, -1, False);  add_1 = None\0A    detach = torch.ops.aten.detach(_softmax)\0A    expand_3 = torch.ops.aten.expand(_softmax, [1, 12, 128, 128]);  _softmax = None\0A    _reshape_alias_2 = torch.ops.aten._reshape_alias(expand_3, [12, 128, 128], [16384, 128, 1]);  expand_3 = None\0A    expand_4 = torch.ops.aten.expand(permute_1, [1, 12, 128, 32]);  permute_1 = None\0A    _reshape_alias_3 = torch.ops.aten._reshape_alias(expand_4, [12, 128, 32], [32, 384, 1]);  expand_4 = None\0A    bmm_1 = torch.ops.aten.bmm(_reshape_alias_2, _reshape_alias_3)\0A    _unsafe_view_1 = torch.ops.aten._unsafe_view(bmm_1, [1, 12, 128, 32]);  bmm_1 = None\0A    permute_3 = torch.ops.aten.permute(_unsafe_view_1, [0, 2, 1, 3]);  _unsafe_view_1 = None\0A    clone = torch.ops.aten.clone(permute_3, memory_format = 0);  permute_3 = None\0A    view_9 = torch.ops.aten.view(clone, [1, 128, 384]);  clone = None\0A    t_3 = torch.ops.aten.t(params_9)\0A    view_10 = torch.ops.aten.view(view_9, [128, 384]);  view_9 = None\0A    addmm_3 = torch.ops.aten.addmm(params_8, view_10, t_3)\0A    view_11 = torch.ops.aten.view(addmm_3, [1, 128, 384]);  addmm_3 = None\0A    add_2 = torch.ops.aten.add(view_11, getitem);  view_11 = getitem = None\0A    native_layer_norm_1 = torch.ops.aten.native_layer_norm(add_2, [384], params_7, params_6, 1e-12)\0A    getitem_3 = native_layer_norm_1[0]\0A    getitem_4 = native_layer_norm_1[1]\0A    getitem_5 = native_layer_norm_1[2];  native_layer_norm_1 = None\0A    t_4 = torch.ops.aten.t(params_17)\0A    view_12 = torch.ops.aten.view(getitem_3, [128, 384])\0A    addmm_4 = torch.ops.aten.addmm(params_16, view_12, t_4)\0A    view_13 = torch.ops.aten.view(addmm_4, [1, 128, 1536]);  addmm_4 = None\0A    gelu = torch.ops.aten.gelu(view_13)\0A    t_5 = torch.ops.aten.t(params_21)\0A    view_14 = torch.ops.aten.view(gelu, [128, 1536]);  gelu = None\0A    addmm_5 = torch.ops.aten.addmm(params_20, view_14, t_5)\0A    view_15 = torch.ops.aten.view(addmm_5, [1, 128, 384]);  addmm_5 = None\0A    add_3 = torch.ops.aten.add(view_15, getitem_3);  view_15 = getitem_3 = None\0A    native_layer_norm_2 = torch.ops.aten.native_layer_norm(add_3, [384], params_19, params_18, 1e-12)\0A    getitem_6 = native_layer_norm_2[0]\0A    getitem_7 = native_layer_norm_2[1]\0A    getitem_8 = native_layer_norm_2[2];  native_layer_norm_2 = None\0A    t_6 = torch.ops.aten.t(params_29)\0A    view_16 = torch.ops.aten.view(getitem_6, [128, 384])\0A    addmm_6 = torch.ops.aten.addmm(params_28, view_16, t_6)\0A    view_17 = torch.ops.aten.view(addmm_6, [1, 128, 384]);  addmm_6 = None\0A    t_7 = torch.ops.aten.t(params_27)\0A    view_18 = torch.ops.aten.view(getitem_6, [128, 384])\0A    addmm_7 = torch.ops.aten.addmm(params_26, view_18, t_7)\0A    view_19 = torch.ops.aten.view(addmm_7, [1, 128, 384]);  addmm_7 = None\0A    view_20 = torch.ops.aten.view(view_19, [1, 128, 12, 32]);  view_19 = None\0A    permute_4 = torch.ops.aten.permute(view_20, [0, 2, 1, 3]);  view_20 = None\0A    t_8 = torch.ops.aten.t(params_31)\0A    view_21 = torch.ops.aten.view(getitem_6, [128, 384])\0A    addmm_8 = torch.ops.aten.addmm(params_30, view_21, t_8)\0A    view_22 = torch.ops.aten.view(addmm_8, [1, 128, 384]);  addmm_8 = None\0A    view_23 = torch.ops.aten.view(view_22, [1, 128, 12, 32]);  view_22 = None\0A    permute_5 = torch.ops.aten.permute(view_23, [0, 2, 1, 3]);  view_23 = None\0A    view_24 = torch.ops.aten.view(view_17, [1, 128, 12, 32]);  view_17 = None\0A    permute_6 = torch.ops.aten.permute(view_24, [0, 2, 1, 3]);  view_24 = None\0A    transpose_1 = torch.ops.aten.transpose(permute_4, -1, -2);  permute_4 = None\0A    expand_5 = torch.ops.aten.expand(permute_6, [1, 12, 128, 32]);  permute_6 = None\0A    _reshape_alias_4 = torch.ops.aten._reshape_alias(expand_5, [12, 128, 32], [32, 384, 1]);  expand_5 = None\0A    expand_6 = torch.ops.aten.expand(transpose_1, [1, 12, 32, 128]);  transpose_1 = None\0A    _reshape_alias_5 = torch.ops.aten._reshape_alias(expand_6, [12, 32, 128], [32, 1, 384]);  expand_6 = None\0A    bmm_2 = torch.ops.aten.bmm(_reshape_alias_4, _reshape_alias_5)\0A    _unsafe_view_2 = torch.ops.aten._unsafe_view(bmm_2, [1, 12, 128, 128]);  bmm_2 = None\0A    div_1 = torch.ops.aten.div(_unsafe_view_2, 5.656854249492381);  _unsafe_view_2 = None\0A    _tensor_constant0_1 = self._tensor_constant0\0A    add_4 = torch.ops.aten.add(div_1, _tensor_constant0_1);  div_1 = _tensor_constant0_1 = None\0A    _softmax_1 = torch.ops.aten._softmax(add_4, -1, False);  add_4 = None\0A    detach_1 = torch.ops.aten.detach(_softmax_1)\0A    expand_7 = torch.ops.aten.expand(_softmax_1, [1, 12, 128, 128]);  _softmax_1 = None\0A    _reshape_alias_6 = torch.ops.aten._reshape_alias(expand_7, [12, 128, 128], [16384, 128, 1]);  expand_7 = None\0A    expand_8 = torch.ops.aten.expand(permute_5, [1, 12, 128, 32]);  permute_5 = None\0A    _reshape_alias_7 = torch.ops.aten._reshape_alias(expand_8, [12, 128, 32], [32, 384, 1]);  expand_8 = None\0A    bmm_3 = torch.ops.aten.bmm(_reshape_alias_6, _reshape_alias_7)\0A    _unsafe_view_3 = torch.ops.aten._unsafe_view(bmm_3, [1, 12, 128, 32]);  bmm_3 = None\0A    permute_7 = torch.ops.aten.permute(_unsafe_view_3, [0, 2, 1, 3]);  _unsafe_view_3 = None\0A    clone_1 = torch.ops.aten.clone(permute_7, memory_format = 0);  permute_7 = None\0A    view_25 = torch.ops.aten.view(clone_1, [1, 128, 384]);  clone_1 = None\0A    t_9 = torch.ops.aten.t(params_25)\0A    view_26 = torch.ops.aten.view(view_25, [128, 384]);  view_25 = None\0A    addmm_9 = torch.ops.aten.addmm(params_24, view_26, t_9)\0A    view_27 = torch.ops.aten.view(addmm_9, [1, 128, 384]);  addmm_9 = None\0A    add_5 = torch.ops.aten.add(view_27, getitem_6);  view_27 = getitem_6 = None\0A    native_layer_norm_3 = torch.ops.aten.native_layer_norm(add_5, [384], params_23, params_22, 1e-12)\0A    getitem_9 = native_layer_norm_3[0]\0A    getitem_10 = native_layer_norm_3[1]\0A    getitem_11 = native_layer_norm_3[2];  native_layer_norm_3 = None\0A    t_10 = torch.ops.aten.t(params_33)\0A    view_28 = torch.ops.aten.view(getitem_9, [128, 384])\0A    addmm_10 = torch.ops.aten.addmm(params_32, view_28, t_10)\0A    view_29 = torch.ops.aten.view(addmm_10, [1, 128, 1536]);  addmm_10 = None\0A    gelu_1 = torch.ops.aten.gelu(view_29)\0A    t_11 = torch.ops.aten.t(params_37)\0A    view_30 = torch.ops.aten.view(gelu_1, [128, 1536]);  gelu_1 = None\0A    addmm_11 = torch.ops.aten.addmm(params_36, view_30, t_11)\0A    view_31 = torch.ops.aten.view(addmm_11, [1, 128, 384]);  addmm_11 = None\0A    add_6 = torch.ops.aten.add(view_31, getitem_9);  view_31 = getitem_9 = None\0A    native_layer_norm_4 = torch.ops.aten.native_layer_norm(add_6, [384], params_35, params_34, 1e-12)\0A    getitem_12 = native_layer_norm_4[0]\0A    getitem_13 = native_layer_norm_4[1]\0A    getitem_14 = native_layer_norm_4[2];  native_layer_norm_4 = None\0A    t_12 = torch.ops.aten.t(params_77)\0A    view_32 = torch.ops.aten.view(getitem_12, [128, 384])\0A    addmm_12 = torch.ops.aten.addmm(params_76, view_32, t_12)\0A    view_33 = torch.ops.aten.view(addmm_12, [1, 128, 384]);  addmm_12 = None\0A    t_13 = torch.ops.aten.t(params_75)\0A    view_34 = torch.ops.aten.view(getitem_12, [128, 384])\0A    addmm_13 = torch.ops.aten.addmm(params_74, view_34, t_13)\0A    view_35 = torch.ops.aten.view(addmm_13, [1, 128, 384]);  addmm_13 = None\0A    view_36 = torch.ops.aten.view(view_35, [1, 128, 12, 32]);  view_35 = None\0A    permute_8 = torch.ops.aten.permute(view_36, [0, 2, 1, 3]);  view_36 = None\0A    t_14 = torch.ops.aten.t(params_79)\0A    view_37 = torch.ops.aten.view(getitem_12, [128, 384])\0A    addmm_14 = torch.ops.aten.addmm(params_78, view_37, t_14)\0A    view_38 = torch.ops.aten.view(addmm_14, [1, 128, 384]);  addmm_14 = None\0A    view_39 = torch.ops.aten.view(view_38, [1, 128, 12, 32]);  view_38 = None\0A    permute_9 = torch.ops.aten.permute(view_39, [0, 2, 1, 3]);  view_39 = None\0A    view_40 = torch.ops.aten.view(view_33, [1, 128, 12, 32]);  view_33 = None\0A    permute_10 = torch.ops.aten.permute(view_40, [0, 2, 1, 3]);  view_40 = None\0A    transpose_2 = torch.ops.aten.transpose(permute_8, -1, -2);  permute_8 = None\0A    expand_9 = torch.ops.aten.expand(permute_10, [1, 12, 128, 32]);  permute_10 = None\0A    _reshape_alias_8 = torch.ops.aten._reshape_alias(expand_9, [12, 128, 32], [32, 384, 1]);  expand_9 = None\0A    expand_10 = torch.ops.aten.expand(transpose_2, [1, 12, 32, 128]);  transpose_2 = None\0A    _reshape_alias_9 = torch.ops.aten._reshape_alias(expand_10, [12, 32, 128], [32, 1, 384]);  expand_10 = None\0A    bmm_4 = torch.ops.aten.bmm(_reshape_alias_8, _reshape_alias_9)\0A    _unsafe_view_4 = torch.ops.aten._unsafe_view(bmm_4, [1, 12, 128, 128]);  bmm_4 = None\0A    div_2 = torch.ops.aten.div(_unsafe_view_4, 5.656854249492381);  _unsafe_view_4 = None\0A    _tensor_constant0_2 = self._tensor_constant0\0A    add_7 = torch.ops.aten.add(div_2, _tensor_constant0_2);  div_2 = _tensor_constant0_2 = None\0A    _softmax_2 = torch.ops.aten._softmax(add_7, -1, False);  add_7 = None\0A    detach_2 = torch.ops.aten.detach(_softmax_2)\0A    expand_11 = torch.ops.aten.expand(_softmax_2, [1, 12, 128, 128]);  _softmax_2 = None\0A    _reshape_alias_10 = torch.ops.aten._reshape_alias(expand_11, [12, 128, 128], [16384, 128, 1]);  expand_11 = None\0A    expand_12 = torch.ops.aten.expand(permute_9, [1, 12, 128, 32]);  permute_9 = None\0A    _reshape_alias_11 = torch.ops.aten._reshape_alias(expand_12, [12, 128, 32], [32, 384, 1]);  expand_12 = None\0A    bmm_5 = torch.ops.aten.bmm(_reshape_alias_10, _reshape_alias_11)\0A    _unsafe_view_5 = torch.ops.aten._unsafe_view(bmm_5, [1, 12, 128, 32]);  bmm_5 = None\0A    permute_11 = torch.ops.aten.permute(_unsafe_view_5, [0, 2, 1, 3]);  _unsafe_view_5 = None\0A    clone_2 = torch.ops.aten.clone(permute_11, memory_format = 0);  permute_11 = None\0A    view_41 = torch.ops.aten.view(clone_2, [1, 128, 384]);  clone_2 = None\0A    t_15 = torch.ops.aten.t(params_73)\0A    view_42 = torch.ops.aten.view(view_41, [128, 384]);  view_41 = None\0A    addmm_15 = torch.ops.aten.addmm(params_72, view_42, t_15)\0A    view_43 = torch.ops.aten.view(addmm_15, [1, 128, 384]);  addmm_15 = None\0A    add_8 = torch.ops.aten.add(view_43, getitem_12);  view_43 = getitem_12 = None\0A    native_layer_norm_5 = torch.ops.aten.native_layer_norm(add_8, [384], params_71, params_70, 1e-12)\0A    getitem_15 = native_layer_norm_5[0]\0A    getitem_16 = native_layer_norm_5[1]\0A    getitem_17 = native_layer_norm_5[2];  native_layer_norm_5 = None\0A    t_16 = torch.ops.aten.t(params_81)\0A    view_44 = torch.ops.aten.view(getitem_15, [128, 384])\0A    addmm_16 = torch.ops.aten.addmm(params_80, view_44, t_16)\0A    view_45 = torch.ops.aten.view(addmm_16, [1, 128, 1536]);  addmm_16 = None\0A    gelu_2 = torch.ops.aten.gelu(view_45)\0A    t_17 = torch.ops.aten.t(params_85)\0A    view_46 = torch.ops.aten.view(gelu_2, [128, 1536]);  gelu_2 = None\0A    addmm_17 = torch.ops.aten.addmm(params_84, view_46, t_17)\0A    view_47 = torch.ops.aten.view(addmm_17, [1, 128, 384]);  addmm_17 = None\0A    add_9 = torch.ops.aten.add(view_47, getitem_15);  view_47 = getitem_15 = None\0A    native_layer_norm_6 = torch.ops.aten.native_layer_norm(add_9, [384], params_83, params_82, 1e-12)\0A    getitem_18 = native_layer_norm_6[0]\0A    getitem_19 = native_layer_norm_6[1]\0A    getitem_20 = native_layer_norm_6[2];  native_layer_norm_6 = None\0A    t_18 = torch.ops.aten.t(params_93)\0A    view_48 = torch.ops.aten.view(getitem_18, [128, 384])\0A    addmm_18 = torch.ops.aten.addmm(params_92, view_48, t_18)\0A    view_49 = torch.ops.aten.view(addmm_18, [1, 128, 384]);  addmm_18 = None\0A    t_19 = torch.ops.aten.t(params_91)\0A    view_50 = torch.ops.aten.view(getitem_18, [128, 384])\0A    addmm_19 = torch.ops.aten.addmm(params_90, view_50, t_19)\0A    view_51 = torch.ops.aten.view(addmm_19, [1, 128, 384]);  addmm_19 = None\0A    view_52 = torch.ops.aten.view(view_51, [1, 128, 12, 32]);  view_51 = None\0A    permute_12 = torch.ops.aten.permute(view_52, [0, 2, 1, 3]);  view_52 = None\0A    t_20 = torch.ops.aten.t(params_95)\0A    view_53 = torch.ops.aten.view(getitem_18, [128, 384])\0A    addmm_20 = torch.ops.aten.addmm(params_94, view_53, t_20)\0A    view_54 = torch.ops.aten.view(addmm_20, [1, 128, 384]);  addmm_20 = None\0A    view_55 = torch.ops.aten.view(view_54, [1, 128, 12, 32]);  view_54 = None\0A    permute_13 = torch.ops.aten.permute(view_55, [0, 2, 1, 3]);  view_55 = None\0A    view_56 = torch.ops.aten.view(view_49, [1, 128, 12, 32]);  view_49 = None\0A    permute_14 = torch.ops.aten.permute(view_56, [0, 2, 1, 3]);  view_56 = None\0A    transpose_3 = torch.ops.aten.transpose(permute_12, -1, -2);  permute_12 = None\0A    expand_13 = torch.ops.aten.expand(permute_14, [1, 12, 128, 32]);  permute_14 = None\0A    _reshape_alias_12 = torch.ops.aten._reshape_alias(expand_13, [12, 128, 32], [32, 384, 1]);  expand_13 = None\0A    expand_14 = torch.ops.aten.expand(transpose_3, [1, 12, 32, 128]);  transpose_3 = None\0A    _reshape_alias_13 = torch.ops.aten._reshape_alias(expand_14, [12, 32, 128], [32, 1, 384]);  expand_14 = None\0A    bmm_6 = torch.ops.aten.bmm(_reshape_alias_12, _reshape_alias_13)\0A    _unsafe_view_6 = torch.ops.aten._unsafe_view(bmm_6, [1, 12, 128, 128]);  bmm_6 = None\0A    div_3 = torch.ops.aten.div(_unsafe_view_6, 5.656854249492381);  _unsafe_view_6 = None\0A    _tensor_constant0_3 = self._tensor_constant0\0A    add_10 = torch.ops.aten.add(div_3, _tensor_constant0_3);  div_3 = _tensor_constant0_3 = None\0A    _softmax_3 = torch.ops.aten._softmax(add_10, -1, False);  add_10 = None\0A    detach_3 = torch.ops.aten.detach(_softmax_3)\0A    expand_15 = torch.ops.aten.expand(_softmax_3, [1, 12, 128, 128]);  _softmax_3 = None\0A    _reshape_alias_14 = torch.ops.aten._reshape_alias(expand_15, [12, 128, 128], [16384, 128, 1]);  expand_15 = None\0A    expand_16 = torch.ops.aten.expand(permute_13, [1, 12, 128, 32]);  permute_13 = None\0A    _reshape_alias_15 = torch.ops.aten._reshape_alias(expand_16, [12, 128, 32], [32, 384, 1]);  expand_16 = None\0A    bmm_7 = torch.ops.aten.bmm(_reshape_alias_14, _reshape_alias_15)\0A    _unsafe_view_7 = torch.ops.aten._unsafe_view(bmm_7, [1, 12, 128, 32]);  bmm_7 = None\0A    permute_15 = torch.ops.aten.permute(_unsafe_view_7, [0, 2, 1, 3]);  _unsafe_view_7 = None\0A    clone_3 = torch.ops.aten.clone(permute_15, memory_format = 0);  permute_15 = None\0A    view_57 = torch.ops.aten.view(clone_3, [1, 128, 384]);  clone_3 = None\0A    t_21 = torch.ops.aten.t(params_89)\0A    view_58 = torch.ops.aten.view(view_57, [128, 384]);  view_57 = None\0A    addmm_21 = torch.ops.aten.addmm(params_88, view_58, t_21)\0A    view_59 = torch.ops.aten.view(addmm_21, [1, 128, 384]);  addmm_21 = None\0A    add_11 = torch.ops.aten.add(view_59, getitem_18);  view_59 = getitem_18 = None\0A    native_layer_norm_7 = torch.ops.aten.native_layer_norm(add_11, [384], params_87, params_86, 1e-12)\0A    getitem_21 = native_layer_norm_7[0]\0A    getitem_22 = native_layer_norm_7[1]\0A    getitem_23 = native_layer_norm_7[2];  native_layer_norm_7 = None\0A    t_22 = torch.ops.aten.t(params_97)\0A    view_60 = torch.ops.aten.view(getitem_21, [128, 384])\0A    addmm_22 = torch.ops.aten.addmm(params_96, view_60, t_22)\0A    view_61 = torch.ops.aten.view(addmm_22, [1, 128, 1536]);  addmm_22 = None\0A    gelu_3 = torch.ops.aten.gelu(view_61)\0A    t_23 = torch.ops.aten.t(params_101)\0A    view_62 = torch.ops.aten.view(gelu_3, [128, 1536]);  gelu_3 = None\0A    addmm_23 = torch.ops.aten.addmm(params_100, view_62, t_23)\0A    view_63 = torch.ops.aten.view(addmm_23, [1, 128, 384]);  addmm_23 = None\0A    add_12 = torch.ops.aten.add(view_63, getitem_21);  view_63 = getitem_21 = None\0A    native_layer_norm_8 = torch.ops.aten.native_layer_norm(add_12, [384], params_99, params_98, 1e-12)\0A    getitem_24 = native_layer_norm_8[0]\0A    getitem_25 = native_layer_norm_8[1]\0A    getitem_26 = native_layer_norm_8[2];  native_layer_norm_8 = None\0A    t_24 = torch.ops.aten.t(params_109)\0A    view_64 = torch.ops.aten.view(getitem_24, [128, 384])\0A    addmm_24 = torch.ops.aten.addmm(params_108, view_64, t_24)\0A    view_65 = torch.ops.aten.view(addmm_24, [1, 128, 384]);  addmm_24 = None\0A    t_25 = torch.ops.aten.t(params_107)\0A    view_66 = torch.ops.aten.view(getitem_24, [128, 384])\0A    addmm_25 = torch.ops.aten.addmm(params_106, view_66, t_25)\0A    view_67 = torch.ops.aten.view(addmm_25, [1, 128, 384]);  addmm_25 = None\0A    view_68 = torch.ops.aten.view(view_67, [1, 128, 12, 32]);  view_67 = None\0A    permute_16 = torch.ops.aten.permute(view_68, [0, 2, 1, 3]);  view_68 = None\0A    t_26 = torch.ops.aten.t(params_111)\0A    view_69 = torch.ops.aten.view(getitem_24, [128, 384])\0A    addmm_26 = torch.ops.aten.addmm(params_110, view_69, t_26)\0A    view_70 = torch.ops.aten.view(addmm_26, [1, 128, 384]);  addmm_26 = None\0A    view_71 = torch.ops.aten.view(view_70, [1, 128, 12, 32]);  view_70 = None\0A    permute_17 = torch.ops.aten.permute(view_71, [0, 2, 1, 3]);  view_71 = None\0A    view_72 = torch.ops.aten.view(view_65, [1, 128, 12, 32]);  view_65 = None\0A    permute_18 = torch.ops.aten.permute(view_72, [0, 2, 1, 3]);  view_72 = None\0A    transpose_4 = torch.ops.aten.transpose(permute_16, -1, -2);  permute_16 = None\0A    expand_17 = torch.ops.aten.expand(permute_18, [1, 12, 128, 32]);  permute_18 = None\0A    _reshape_alias_16 = torch.ops.aten._reshape_alias(expand_17, [12, 128, 32], [32, 384, 1]);  expand_17 = None\0A    expand_18 = torch.ops.aten.expand(transpose_4, [1, 12, 32, 128]);  transpose_4 = None\0A    _reshape_alias_17 = torch.ops.aten._reshape_alias(expand_18, [12, 32, 128], [32, 1, 384]);  expand_18 = None\0A    bmm_8 = torch.ops.aten.bmm(_reshape_alias_16, _reshape_alias_17)\0A    _unsafe_view_8 = torch.ops.aten._unsafe_view(bmm_8, [1, 12, 128, 128]);  bmm_8 = None\0A    div_4 = torch.ops.aten.div(_unsafe_view_8, 5.656854249492381);  _unsafe_view_8 = None\0A    _tensor_constant0_4 = self._tensor_constant0\0A    add_13 = torch.ops.aten.add(div_4, _tensor_constant0_4);  div_4 = _tensor_constant0_4 = None\0A    _softmax_4 = torch.ops.aten._softmax(add_13, -1, False);  add_13 = None\0A    detach_4 = torch.ops.aten.detach(_softmax_4)\0A    expand_19 = torch.ops.aten.expand(_softmax_4, [1, 12, 128, 128]);  _softmax_4 = None\0A    _reshape_alias_18 = torch.ops.aten._reshape_alias(expand_19, [12, 128, 128], [16384, 128, 1]);  expand_19 = None\0A    expand_20 = torch.ops.aten.expand(permute_17, [1, 12, 128, 32]);  permute_17 = None\0A    _reshape_alias_19 = torch.ops.aten._reshape_alias(expand_20, [12, 128, 32], [32, 384, 1]);  expand_20 = None\0A    bmm_9 = torch.ops.aten.bmm(_reshape_alias_18, _reshape_alias_19)\0A    _unsafe_view_9 = torch.ops.aten._unsafe_view(bmm_9, [1, 12, 128, 32]);  bmm_9 = None\0A    permute_19 = torch.ops.aten.permute(_unsafe_view_9, [0, 2, 1, 3]);  _unsafe_view_9 = None\0A    clone_4 = torch.ops.aten.clone(permute_19, memory_format = 0);  permute_19 = None\0A    view_73 = torch.ops.aten.view(clone_4, [1, 128, 384]);  clone_4 = None\0A    t_27 = torch.ops.aten.t(params_105)\0A    view_74 = torch.ops.aten.view(view_73, [128, 384]);  view_73 = None\0A    addmm_27 = torch.ops.aten.addmm(params_104, view_74, t_27)\0A    view_75 = torch.ops.aten.view(addmm_27, [1, 128, 384]);  addmm_27 = None\0A    add_14 = torch.ops.aten.add(view_75, getitem_24);  view_75 = getitem_24 = None\0A    native_layer_norm_9 = torch.ops.aten.native_layer_norm(add_14, [384], params_103, params_102, 1e-12)\0A    getitem_27 = native_layer_norm_9[0]\0A    getitem_28 = native_layer_norm_9[1]\0A    getitem_29 = native_layer_norm_9[2];  native_layer_norm_9 = None\0A    t_28 = torch.ops.aten.t(params_113)\0A    view_76 = torch.ops.aten.view(getitem_27, [128, 384])\0A    addmm_28 = torch.ops.aten.addmm(params_112, view_76, t_28)\0A    view_77 = torch.ops.aten.view(addmm_28, [1, 128, 1536]);  addmm_28 = None\0A    gelu_4 = torch.ops.aten.gelu(view_77)\0A    t_29 = torch.ops.aten.t(params_117)\0A    view_78 = torch.ops.aten.view(gelu_4, [128, 1536]);  gelu_4 = None\0A    addmm_29 = torch.ops.aten.addmm(params_116, view_78, t_29)\0A    view_79 = torch.ops.aten.view(addmm_29, [1, 128, 384]);  addmm_29 = None\0A    add_15 = torch.ops.aten.add(view_79, getitem_27);  view_79 = getitem_27 = None\0A    native_layer_norm_10 = torch.ops.aten.native_layer_norm(add_15, [384], params_115, params_114, 1e-12)\0A    getitem_30 = native_layer_norm_10[0]\0A    getitem_31 = native_layer_norm_10[1]\0A    getitem_32 = native_layer_norm_10[2];  native_layer_norm_10 = None\0A    t_30 = torch.ops.aten.t(params_125)\0A    view_80 = torch.ops.aten.view(getitem_30, [128, 384])\0A    addmm_30 = torch.ops.aten.addmm(params_124, view_80, t_30)\0A    view_81 = torch.ops.aten.view(addmm_30, [1, 128, 384]);  addmm_30 = None\0A    t_31 = torch.ops.aten.t(params_123)\0A    view_82 = torch.ops.aten.view(getitem_30, [128, 384])\0A    addmm_31 = torch.ops.aten.addmm(params_122, view_82, t_31)\0A    view_83 = torch.ops.aten.view(addmm_31, [1, 128, 384]);  addmm_31 = None\0A    view_84 = torch.ops.aten.view(view_83, [1, 128, 12, 32]);  view_83 = None\0A    permute_20 = torch.ops.aten.permute(view_84, [0, 2, 1, 3]);  view_84 = None\0A    t_32 = torch.ops.aten.t(params_127)\0A    view_85 = torch.ops.aten.view(getitem_30, [128, 384])\0A    addmm_32 = torch.ops.aten.addmm(params_126, view_85, t_32)\0A    view_86 = torch.ops.aten.view(addmm_32, [1, 128, 384]);  addmm_32 = None\0A    view_87 = torch.ops.aten.view(view_86, [1, 128, 12, 32]);  view_86 = None\0A    permute_21 = torch.ops.aten.permute(view_87, [0, 2, 1, 3]);  view_87 = None\0A    view_88 = torch.ops.aten.view(view_81, [1, 128, 12, 32]);  view_81 = None\0A    permute_22 = torch.ops.aten.permute(view_88, [0, 2, 1, 3]);  view_88 = None\0A    transpose_5 = torch.ops.aten.transpose(permute_20, -1, -2);  permute_20 = None\0A    expand_21 = torch.ops.aten.expand(permute_22, [1, 12, 128, 32]);  permute_22 = None\0A    _reshape_alias_20 = torch.ops.aten._reshape_alias(expand_21, [12, 128, 32], [32, 384, 1]);  expand_21 = None\0A    expand_22 = torch.ops.aten.expand(transpose_5, [1, 12, 32, 128]);  transpose_5 = None\0A    _reshape_alias_21 = torch.ops.aten._reshape_alias(expand_22, [12, 32, 128], [32, 1, 384]);  expand_22 = None\0A    bmm_10 = torch.ops.aten.bmm(_reshape_alias_20, _reshape_alias_21)\0A    _unsafe_view_10 = torch.ops.aten._unsafe_view(bmm_10, [1, 12, 128, 128]);  bmm_10 = None\0A    div_5 = torch.ops.aten.div(_unsafe_view_10, 5.656854249492381);  _unsafe_view_10 = None\0A    _tensor_constant0_5 = self._tensor_constant0\0A    add_16 = torch.ops.aten.add(div_5, _tensor_constant0_5);  div_5 = _tensor_constant0_5 = None\0A    _softmax_5 = torch.ops.aten._softmax(add_16, -1, False);  add_16 = None\0A    detach_5 = torch.ops.aten.detach(_softmax_5)\0A    expand_23 = torch.ops.aten.expand(_softmax_5, [1, 12, 128, 128]);  _softmax_5 = None\0A    _reshape_alias_22 = torch.ops.aten._reshape_alias(expand_23, [12, 128, 128], [16384, 128, 1]);  expand_23 = None\0A    expand_24 = torch.ops.aten.expand(permute_21, [1, 12, 128, 32]);  permute_21 = None\0A    _reshape_alias_23 = torch.ops.aten._reshape_alias(expand_24, [12, 128, 32], [32, 384, 1]);  expand_24 = None\0A    bmm_11 = torch.ops.aten.bmm(_reshape_alias_22, _reshape_alias_23)\0A    _unsafe_view_11 = torch.ops.aten._unsafe_view(bmm_11, [1, 12, 128, 32]);  bmm_11 = None\0A    permute_23 = torch.ops.aten.permute(_unsafe_view_11, [0, 2, 1, 3]);  _unsafe_view_11 = None\0A    clone_5 = torch.ops.aten.clone(permute_23, memory_format = 0);  permute_23 = None\0A    view_89 = torch.ops.aten.view(clone_5, [1, 128, 384]);  clone_5 = None\0A    t_33 = torch.ops.aten.t(params_121)\0A    view_90 = torch.ops.aten.view(view_89, [128, 384]);  view_89 = None\0A    addmm_33 = torch.ops.aten.addmm(params_120, view_90, t_33)\0A    view_91 = torch.ops.aten.view(addmm_33, [1, 128, 384]);  addmm_33 = None\0A    add_17 = torch.ops.aten.add(view_91, getitem_30);  view_91 = getitem_30 = None\0A    native_layer_norm_11 = torch.ops.aten.native_layer_norm(add_17, [384], params_119, params_118, 1e-12)\0A    getitem_33 = native_layer_norm_11[0]\0A    getitem_34 = native_layer_norm_11[1]\0A    getitem_35 = native_layer_norm_11[2];  native_layer_norm_11 = None\0A    t_34 = torch.ops.aten.t(params_129)\0A    view_92 = torch.ops.aten.view(getitem_33, [128, 384])\0A    addmm_34 = torch.ops.aten.addmm(params_128, view_92, t_34)\0A    view_93 = torch.ops.aten.view(addmm_34, [1, 128, 1536]);  addmm_34 = None\0A    gelu_5 = torch.ops.aten.gelu(view_93)\0A    t_35 = torch.ops.aten.t(params_133)\0A    view_94 = torch.ops.aten.view(gelu_5, [128, 1536]);  gelu_5 = None\0A    addmm_35 = torch.ops.aten.addmm(params_132, view_94, t_35)\0A    view_95 = torch.ops.aten.view(addmm_35, [1, 128, 384]);  addmm_35 = None\0A    add_18 = torch.ops.aten.add(view_95, getitem_33);  view_95 = getitem_33 = None\0A    native_layer_norm_12 = torch.ops.aten.native_layer_norm(add_18, [384], params_131, params_130, 1e-12)\0A    getitem_36 = native_layer_norm_12[0]\0A    getitem_37 = native_layer_norm_12[1]\0A    getitem_38 = native_layer_norm_12[2];  native_layer_norm_12 = None\0A    t_36 = torch.ops.aten.t(params_141)\0A    view_96 = torch.ops.aten.view(getitem_36, [128, 384])\0A    addmm_36 = torch.ops.aten.addmm(params_140, view_96, t_36)\0A    view_97 = torch.ops.aten.view(addmm_36, [1, 128, 384]);  addmm_36 = None\0A    t_37 = torch.ops.aten.t(params_139)\0A    view_98 = torch.ops.aten.view(getitem_36, [128, 384])\0A    addmm_37 = torch.ops.aten.addmm(params_138, view_98, t_37)\0A    view_99 = torch.ops.aten.view(addmm_37, [1, 128, 384]);  addmm_37 = None\0A    view_100 = torch.ops.aten.view(view_99, [1, 128, 12, 32]);  view_99 = None\0A    permute_24 = torch.ops.aten.permute(view_100, [0, 2, 1, 3]);  view_100 = None\0A    t_38 = torch.ops.aten.t(params_143)\0A    view_101 = torch.ops.aten.view(getitem_36, [128, 384])\0A    addmm_38 = torch.ops.aten.addmm(params_142, view_101, t_38)\0A    view_102 = torch.ops.aten.view(addmm_38, [1, 128, 384]);  addmm_38 = None\0A    view_103 = torch.ops.aten.view(view_102, [1, 128, 12, 32]);  view_102 = None\0A    permute_25 = torch.ops.aten.permute(view_103, [0, 2, 1, 3]);  view_103 = None\0A    view_104 = torch.ops.aten.view(view_97, [1, 128, 12, 32]);  view_97 = None\0A    permute_26 = torch.ops.aten.permute(view_104, [0, 2, 1, 3]);  view_104 = None\0A    transpose_6 = torch.ops.aten.transpose(permute_24, -1, -2);  permute_24 = None\0A    expand_25 = torch.ops.aten.expand(permute_26, [1, 12, 128, 32]);  permute_26 = None\0A    _reshape_alias_24 = torch.ops.aten._reshape_alias(expand_25, [12, 128, 32], [32, 384, 1]);  expand_25 = None\0A    expand_26 = torch.ops.aten.expand(transpose_6, [1, 12, 32, 128]);  transpose_6 = None\0A    _reshape_alias_25 = torch.ops.aten._reshape_alias(expand_26, [12, 32, 128], [32, 1, 384]);  expand_26 = None\0A    bmm_12 = torch.ops.aten.bmm(_reshape_alias_24, _reshape_alias_25)\0A    _unsafe_view_12 = torch.ops.aten._unsafe_view(bmm_12, [1, 12, 128, 128]);  bmm_12 = None\0A    div_6 = torch.ops.aten.div(_unsafe_view_12, 5.656854249492381);  _unsafe_view_12 = None\0A    _tensor_constant0_6 = self._tensor_constant0\0A    add_19 = torch.ops.aten.add(div_6, _tensor_constant0_6);  div_6 = _tensor_constant0_6 = None\0A    _softmax_6 = torch.ops.aten._softmax(add_19, -1, False);  add_19 = None\0A    detach_6 = torch.ops.aten.detach(_softmax_6)\0A    expand_27 = torch.ops.aten.expand(_softmax_6, [1, 12, 128, 128]);  _softmax_6 = None\0A    _reshape_alias_26 = torch.ops.aten._reshape_alias(expand_27, [12, 128, 128], [16384, 128, 1]);  expand_27 = None\0A    expand_28 = torch.ops.aten.expand(permute_25, [1, 12, 128, 32]);  permute_25 = None\0A    _reshape_alias_27 = torch.ops.aten._reshape_alias(expand_28, [12, 128, 32], [32, 384, 1]);  expand_28 = None\0A    bmm_13 = torch.ops.aten.bmm(_reshape_alias_26, _reshape_alias_27)\0A    _unsafe_view_13 = torch.ops.aten._unsafe_view(bmm_13, [1, 12, 128, 32]);  bmm_13 = None\0A    permute_27 = torch.ops.aten.permute(_unsafe_view_13, [0, 2, 1, 3]);  _unsafe_view_13 = None\0A    clone_6 = torch.ops.aten.clone(permute_27, memory_format = 0);  permute_27 = None\0A    view_105 = torch.ops.aten.view(clone_6, [1, 128, 384]);  clone_6 = None\0A    t_39 = torch.ops.aten.t(params_137)\0A    view_106 = torch.ops.aten.view(view_105, [128, 384]);  view_105 = None\0A    addmm_39 = torch.ops.aten.addmm(params_136, view_106, t_39)\0A    view_107 = torch.ops.aten.view(addmm_39, [1, 128, 384]);  addmm_39 = None\0A    add_20 = torch.ops.aten.add(view_107, getitem_36);  view_107 = getitem_36 = None\0A    native_layer_norm_13 = torch.ops.aten.native_layer_norm(add_20, [384], params_135, params_134, 1e-12)\0A    getitem_39 = native_layer_norm_13[0]\0A    getitem_40 = native_layer_norm_13[1]\0A    getitem_41 = native_layer_norm_13[2];  native_layer_norm_13 = None\0A    t_40 = torch.ops.aten.t(params_145)\0A    view_108 = torch.ops.aten.view(getitem_39, [128, 384])\0A    addmm_40 = torch.ops.aten.addmm(params_144, view_108, t_40)\0A    view_109 = torch.ops.aten.view(addmm_40, [1, 128, 1536]);  addmm_40 = None\0A    gelu_6 = torch.ops.aten.gelu(view_109)\0A    t_41 = torch.ops.aten.t(params_149)\0A    view_110 = torch.ops.aten.view(gelu_6, [128, 1536]);  gelu_6 = None\0A    addmm_41 = torch.ops.aten.addmm(params_148, view_110, t_41)\0A    view_111 = torch.ops.aten.view(addmm_41, [1, 128, 384]);  addmm_41 = None\0A    add_21 = torch.ops.aten.add(view_111, getitem_39);  view_111 = getitem_39 = None\0A    native_layer_norm_14 = torch.ops.aten.native_layer_norm(add_21, [384], params_147, params_146, 1e-12)\0A    getitem_42 = native_layer_norm_14[0]\0A    getitem_43 = native_layer_norm_14[1]\0A    getitem_44 = native_layer_norm_14[2];  native_layer_norm_14 = None\0A    t_42 = torch.ops.aten.t(params_157)\0A    view_112 = torch.ops.aten.view(getitem_42, [128, 384])\0A    addmm_42 = torch.ops.aten.addmm(params_156, view_112, t_42)\0A    view_113 = torch.ops.aten.view(addmm_42, [1, 128, 384]);  addmm_42 = None\0A    t_43 = torch.ops.aten.t(params_155)\0A    view_114 = torch.ops.aten.view(getitem_42, [128, 384])\0A    addmm_43 = torch.ops.aten.addmm(params_154, view_114, t_43)\0A    view_115 = torch.ops.aten.view(addmm_43, [1, 128, 384]);  addmm_43 = None\0A    view_116 = torch.ops.aten.view(view_115, [1, 128, 12, 32]);  view_115 = None\0A    permute_28 = torch.ops.aten.permute(view_116, [0, 2, 1, 3]);  view_116 = None\0A    t_44 = torch.ops.aten.t(params_159)\0A    view_117 = torch.ops.aten.view(getitem_42, [128, 384])\0A    addmm_44 = torch.ops.aten.addmm(params_158, view_117, t_44)\0A    view_118 = torch.ops.aten.view(addmm_44, [1, 128, 384]);  addmm_44 = None\0A    view_119 = torch.ops.aten.view(view_118, [1, 128, 12, 32]);  view_118 = None\0A    permute_29 = torch.ops.aten.permute(view_119, [0, 2, 1, 3]);  view_119 = None\0A    view_120 = torch.ops.aten.view(view_113, [1, 128, 12, 32]);  view_113 = None\0A    permute_30 = torch.ops.aten.permute(view_120, [0, 2, 1, 3]);  view_120 = None\0A    transpose_7 = torch.ops.aten.transpose(permute_28, -1, -2);  permute_28 = None\0A    expand_29 = torch.ops.aten.expand(permute_30, [1, 12, 128, 32]);  permute_30 = None\0A    _reshape_alias_28 = torch.ops.aten._reshape_alias(expand_29, [12, 128, 32], [32, 384, 1]);  expand_29 = None\0A    expand_30 = torch.ops.aten.expand(transpose_7, [1, 12, 32, 128]);  transpose_7 = None\0A    _reshape_alias_29 = torch.ops.aten._reshape_alias(expand_30, [12, 32, 128], [32, 1, 384]);  expand_30 = None\0A    bmm_14 = torch.ops.aten.bmm(_reshape_alias_28, _reshape_alias_29)\0A    _unsafe_view_14 = torch.ops.aten._unsafe_view(bmm_14, [1, 12, 128, 128]);  bmm_14 = None\0A    div_7 = torch.ops.aten.div(_unsafe_view_14, 5.656854249492381);  _unsafe_view_14 = None\0A    _tensor_constant0_7 = self._tensor_constant0\0A    add_22 = torch.ops.aten.add(div_7, _tensor_constant0_7);  div_7 = _tensor_constant0_7 = None\0A    _softmax_7 = torch.ops.aten._softmax(add_22, -1, False);  add_22 = None\0A    detach_7 = torch.ops.aten.detach(_softmax_7)\0A    expand_31 = torch.ops.aten.expand(_softmax_7, [1, 12, 128, 128]);  _softmax_7 = None\0A    _reshape_alias_30 = torch.ops.aten._reshape_alias(expand_31, [12, 128, 128], [16384, 128, 1]);  expand_31 = None\0A    expand_32 = torch.ops.aten.expand(permute_29, [1, 12, 128, 32]);  permute_29 = None\0A    _reshape_alias_31 = torch.ops.aten._reshape_alias(expand_32, [12, 128, 32], [32, 384, 1]);  expand_32 = None\0A    bmm_15 = torch.ops.aten.bmm(_reshape_alias_30, _reshape_alias_31)\0A    _unsafe_view_15 = torch.ops.aten._unsafe_view(bmm_15, [1, 12, 128, 32]);  bmm_15 = None\0A    permute_31 = torch.ops.aten.permute(_unsafe_view_15, [0, 2, 1, 3]);  _unsafe_view_15 = None\0A    clone_7 = torch.ops.aten.clone(permute_31, memory_format = 0);  permute_31 = None\0A    view_121 = torch.ops.aten.view(clone_7, [1, 128, 384]);  clone_7 = None\0A    t_45 = torch.ops.aten.t(params_153)\0A    view_122 = torch.ops.aten.view(view_121, [128, 384]);  view_121 = None\0A    addmm_45 = torch.ops.aten.addmm(params_152, view_122, t_45)\0A    view_123 = torch.ops.aten.view(addmm_45, [1, 128, 384]);  addmm_45 = None\0A    add_23 = torch.ops.aten.add(view_123, getitem_42);  view_123 = getitem_42 = None\0A    native_layer_norm_15 = torch.ops.aten.native_layer_norm(add_23, [384], params_151, params_150, 1e-12)\0A    getitem_45 = native_layer_norm_15[0]\0A    getitem_46 = native_layer_norm_15[1]\0A    getitem_47 = native_layer_norm_15[2];  native_layer_norm_15 = None\0A    t_46 = torch.ops.aten.t(params_161)\0A    view_124 = torch.ops.aten.view(getitem_45, [128, 384])\0A    addmm_46 = torch.ops.aten.addmm(params_160, view_124, t_46)\0A    view_125 = torch.ops.aten.view(addmm_46, [1, 128, 1536]);  addmm_46 = None\0A    gelu_7 = torch.ops.aten.gelu(view_125)\0A    t_47 = torch.ops.aten.t(params_165)\0A    view_126 = torch.ops.aten.view(gelu_7, [128, 1536]);  gelu_7 = None\0A    addmm_47 = torch.ops.aten.addmm(params_164, view_126, t_47)\0A    view_127 = torch.ops.aten.view(addmm_47, [1, 128, 384]);  addmm_47 = None\0A    add_24 = torch.ops.aten.add(view_127, getitem_45);  view_127 = getitem_45 = None\0A    native_layer_norm_16 = torch.ops.aten.native_layer_norm(add_24, [384], params_163, params_162, 1e-12)\0A    getitem_48 = native_layer_norm_16[0]\0A    getitem_49 = native_layer_norm_16[1]\0A    getitem_50 = native_layer_norm_16[2];  native_layer_norm_16 = None\0A    t_48 = torch.ops.aten.t(params_173)\0A    view_128 = torch.ops.aten.view(getitem_48, [128, 384])\0A    addmm_48 = torch.ops.aten.addmm(params_172, view_128, t_48)\0A    view_129 = torch.ops.aten.view(addmm_48, [1, 128, 384]);  addmm_48 = None\0A    t_49 = torch.ops.aten.t(params_171)\0A    view_130 = torch.ops.aten.view(getitem_48, [128, 384])\0A    addmm_49 = torch.ops.aten.addmm(params_170, view_130, t_49)\0A    view_131 = torch.ops.aten.view(addmm_49, [1, 128, 384]);  addmm_49 = None\0A    view_132 = torch.ops.aten.view(view_131, [1, 128, 12, 32]);  view_131 = None\0A    permute_32 = torch.ops.aten.permute(view_132, [0, 2, 1, 3]);  view_132 = None\0A    t_50 = torch.ops.aten.t(params_175)\0A    view_133 = torch.ops.aten.view(getitem_48, [128, 384])\0A    addmm_50 = torch.ops.aten.addmm(params_174, view_133, t_50)\0A    view_134 = torch.ops.aten.view(addmm_50, [1, 128, 384]);  addmm_50 = None\0A    view_135 = torch.ops.aten.view(view_134, [1, 128, 12, 32]);  view_134 = None\0A    permute_33 = torch.ops.aten.permute(view_135, [0, 2, 1, 3]);  view_135 = None\0A    view_136 = torch.ops.aten.view(view_129, [1, 128, 12, 32]);  view_129 = None\0A    permute_34 = torch.ops.aten.permute(view_136, [0, 2, 1, 3]);  view_136 = None\0A    transpose_8 = torch.ops.aten.transpose(permute_32, -1, -2);  permute_32 = None\0A    expand_33 = torch.ops.aten.expand(permute_34, [1, 12, 128, 32]);  permute_34 = None\0A    _reshape_alias_32 = torch.ops.aten._reshape_alias(expand_33, [12, 128, 32], [32, 384, 1]);  expand_33 = None\0A    expand_34 = torch.ops.aten.expand(transpose_8, [1, 12, 32, 128]);  transpose_8 = None\0A    _reshape_alias_33 = torch.ops.aten._reshape_alias(expand_34, [12, 32, 128], [32, 1, 384]);  expand_34 = None\0A    bmm_16 = torch.ops.aten.bmm(_reshape_alias_32, _reshape_alias_33)\0A    _unsafe_view_16 = torch.ops.aten._unsafe_view(bmm_16, [1, 12, 128, 128]);  bmm_16 = None\0A    div_8 = torch.ops.aten.div(_unsafe_view_16, 5.656854249492381);  _unsafe_view_16 = None\0A    _tensor_constant0_8 = self._tensor_constant0\0A    add_25 = torch.ops.aten.add(div_8, _tensor_constant0_8);  div_8 = _tensor_constant0_8 = None\0A    _softmax_8 = torch.ops.aten._softmax(add_25, -1, False);  add_25 = None\0A    detach_8 = torch.ops.aten.detach(_softmax_8)\0A    expand_35 = torch.ops.aten.expand(_softmax_8, [1, 12, 128, 128]);  _softmax_8 = None\0A    _reshape_alias_34 = torch.ops.aten._reshape_alias(expand_35, [12, 128, 128], [16384, 128, 1]);  expand_35 = None\0A    expand_36 = torch.ops.aten.expand(permute_33, [1, 12, 128, 32]);  permute_33 = None\0A    _reshape_alias_35 = torch.ops.aten._reshape_alias(expand_36, [12, 128, 32], [32, 384, 1]);  expand_36 = None\0A    bmm_17 = torch.ops.aten.bmm(_reshape_alias_34, _reshape_alias_35)\0A    _unsafe_view_17 = torch.ops.aten._unsafe_view(bmm_17, [1, 12, 128, 32]);  bmm_17 = None\0A    permute_35 = torch.ops.aten.permute(_unsafe_view_17, [0, 2, 1, 3]);  _unsafe_view_17 = None\0A    clone_8 = torch.ops.aten.clone(permute_35, memory_format = 0);  permute_35 = None\0A    view_137 = torch.ops.aten.view(clone_8, [1, 128, 384]);  clone_8 = None\0A    t_51 = torch.ops.aten.t(params_169)\0A    view_138 = torch.ops.aten.view(view_137, [128, 384]);  view_137 = None\0A    addmm_51 = torch.ops.aten.addmm(params_168, view_138, t_51)\0A    view_139 = torch.ops.aten.view(addmm_51, [1, 128, 384]);  addmm_51 = None\0A    add_26 = torch.ops.aten.add(view_139, getitem_48);  view_139 = getitem_48 = None\0A    native_layer_norm_17 = torch.ops.aten.native_layer_norm(add_26, [384], params_167, params_166, 1e-12)\0A    getitem_51 = native_layer_norm_17[0]\0A    getitem_52 = native_layer_norm_17[1]\0A    getitem_53 = native_layer_norm_17[2];  native_layer_norm_17 = None\0A    t_52 = torch.ops.aten.t(params_177)\0A    view_140 = torch.ops.aten.view(getitem_51, [128, 384])\0A    addmm_52 = torch.ops.aten.addmm(params_176, view_140, t_52)\0A    view_141 = torch.ops.aten.view(addmm_52, [1, 128, 1536]);  addmm_52 = None\0A    gelu_8 = torch.ops.aten.gelu(view_141)\0A    t_53 = torch.ops.aten.t(params_181)\0A    view_142 = torch.ops.aten.view(gelu_8, [128, 1536]);  gelu_8 = None\0A    addmm_53 = torch.ops.aten.addmm(params_180, view_142, t_53)\0A    view_143 = torch.ops.aten.view(addmm_53, [1, 128, 384]);  addmm_53 = None\0A    add_27 = torch.ops.aten.add(view_143, getitem_51);  view_143 = getitem_51 = None\0A    native_layer_norm_18 = torch.ops.aten.native_layer_norm(add_27, [384], params_179, params_178, 1e-12)\0A    getitem_54 = native_layer_norm_18[0]\0A    getitem_55 = native_layer_norm_18[1]\0A    getitem_56 = native_layer_norm_18[2];  native_layer_norm_18 = None\0A    t_54 = torch.ops.aten.t(params_189)\0A    view_144 = torch.ops.aten.view(getitem_54, [128, 384])\0A    addmm_54 = torch.ops.aten.addmm(params_188, view_144, t_54)\0A    view_145 = torch.ops.aten.view(addmm_54, [1, 128, 384]);  addmm_54 = None\0A    t_55 = torch.ops.aten.t(params_187)\0A    view_146 = torch.ops.aten.view(getitem_54, [128, 384])\0A    addmm_55 = torch.ops.aten.addmm(params_186, view_146, t_55)\0A    view_147 = torch.ops.aten.view(addmm_55, [1, 128, 384]);  addmm_55 = None\0A    view_148 = torch.ops.aten.view(view_147, [1, 128, 12, 32]);  view_147 = None\0A    permute_36 = torch.ops.aten.permute(view_148, [0, 2, 1, 3]);  view_148 = None\0A    t_56 = torch.ops.aten.t(params_191)\0A    view_149 = torch.ops.aten.view(getitem_54, [128, 384])\0A    addmm_56 = torch.ops.aten.addmm(params_190, view_149, t_56)\0A    view_150 = torch.ops.aten.view(addmm_56, [1, 128, 384]);  addmm_56 = None\0A    view_151 = torch.ops.aten.view(view_150, [1, 128, 12, 32]);  view_150 = None\0A    permute_37 = torch.ops.aten.permute(view_151, [0, 2, 1, 3]);  view_151 = None\0A    view_152 = torch.ops.aten.view(view_145, [1, 128, 12, 32]);  view_145 = None\0A    permute_38 = torch.ops.aten.permute(view_152, [0, 2, 1, 3]);  view_152 = None\0A    transpose_9 = torch.ops.aten.transpose(permute_36, -1, -2);  permute_36 = None\0A    expand_37 = torch.ops.aten.expand(permute_38, [1, 12, 128, 32]);  permute_38 = None\0A    _reshape_alias_36 = torch.ops.aten._reshape_alias(expand_37, [12, 128, 32], [32, 384, 1]);  expand_37 = None\0A    expand_38 = torch.ops.aten.expand(transpose_9, [1, 12, 32, 128]);  transpose_9 = None\0A    _reshape_alias_37 = torch.ops.aten._reshape_alias(expand_38, [12, 32, 128], [32, 1, 384]);  expand_38 = None\0A    bmm_18 = torch.ops.aten.bmm(_reshape_alias_36, _reshape_alias_37)\0A    _unsafe_view_18 = torch.ops.aten._unsafe_view(bmm_18, [1, 12, 128, 128]);  bmm_18 = None\0A    div_9 = torch.ops.aten.div(_unsafe_view_18, 5.656854249492381);  _unsafe_view_18 = None\0A    _tensor_constant0_9 = self._tensor_constant0\0A    add_28 = torch.ops.aten.add(div_9, _tensor_constant0_9);  div_9 = _tensor_constant0_9 = None\0A    _softmax_9 = torch.ops.aten._softmax(add_28, -1, False);  add_28 = None\0A    detach_9 = torch.ops.aten.detach(_softmax_9)\0A    expand_39 = torch.ops.aten.expand(_softmax_9, [1, 12, 128, 128]);  _softmax_9 = None\0A    _reshape_alias_38 = torch.ops.aten._reshape_alias(expand_39, [12, 128, 128], [16384, 128, 1]);  expand_39 = None\0A    expand_40 = torch.ops.aten.expand(permute_37, [1, 12, 128, 32]);  permute_37 = None\0A    _reshape_alias_39 = torch.ops.aten._reshape_alias(expand_40, [12, 128, 32], [32, 384, 1]);  expand_40 = None\0A    bmm_19 = torch.ops.aten.bmm(_reshape_alias_38, _reshape_alias_39)\0A    _unsafe_view_19 = torch.ops.aten._unsafe_view(bmm_19, [1, 12, 128, 32]);  bmm_19 = None\0A    permute_39 = torch.ops.aten.permute(_unsafe_view_19, [0, 2, 1, 3]);  _unsafe_view_19 = None\0A    clone_9 = torch.ops.aten.clone(permute_39, memory_format = 0);  permute_39 = None\0A    view_153 = torch.ops.aten.view(clone_9, [1, 128, 384]);  clone_9 = None\0A    t_57 = torch.ops.aten.t(params_185)\0A    view_154 = torch.ops.aten.view(view_153, [128, 384]);  view_153 = None\0A    addmm_57 = torch.ops.aten.addmm(params_184, view_154, t_57)\0A    view_155 = torch.ops.aten.view(addmm_57, [1, 128, 384]);  addmm_57 = None\0A    add_29 = torch.ops.aten.add(view_155, getitem_54);  view_155 = getitem_54 = None\0A    native_layer_norm_19 = torch.ops.aten.native_layer_norm(add_29, [384], params_183, params_182, 1e-12)\0A    getitem_57 = native_layer_norm_19[0]\0A    getitem_58 = native_layer_norm_19[1]\0A    getitem_59 = native_layer_norm_19[2];  native_layer_norm_19 = None\0A    t_58 = torch.ops.aten.t(params_193)\0A    view_156 = torch.ops.aten.view(getitem_57, [128, 384])\0A    addmm_58 = torch.ops.aten.addmm(params_192, view_156, t_58)\0A    view_157 = torch.ops.aten.view(addmm_58, [1, 128, 1536]);  addmm_58 = None\0A    gelu_9 = torch.ops.aten.gelu(view_157)\0A    t_59 = torch.ops.aten.t(params_197)\0A    view_158 = torch.ops.aten.view(gelu_9, [128, 1536]);  gelu_9 = None\0A    addmm_59 = torch.ops.aten.addmm(params_196, view_158, t_59)\0A    view_159 = torch.ops.aten.view(addmm_59, [1, 128, 384]);  addmm_59 = None\0A    add_30 = torch.ops.aten.add(view_159, getitem_57);  view_159 = getitem_57 = None\0A    native_layer_norm_20 = torch.ops.aten.native_layer_norm(add_30, [384], params_195, params_194, 1e-12)\0A    getitem_60 = native_layer_norm_20[0]\0A    getitem_61 = native_layer_norm_20[1]\0A    getitem_62 = native_layer_norm_20[2];  native_layer_norm_20 = None\0A    t_60 = torch.ops.aten.t(params_45)\0A    view_160 = torch.ops.aten.view(getitem_60, [128, 384])\0A    addmm_60 = torch.ops.aten.addmm(params_44, view_160, t_60)\0A    view_161 = torch.ops.aten.view(addmm_60, [1, 128, 384]);  addmm_60 = None\0A    t_61 = torch.ops.aten.t(params_43)\0A    view_162 = torch.ops.aten.view(getitem_60, [128, 384])\0A    addmm_61 = torch.ops.aten.addmm(params_42, view_162, t_61)\0A    view_163 = torch.ops.aten.view(addmm_61, [1, 128, 384]);  addmm_61 = None\0A    view_164 = torch.ops.aten.view(view_163, [1, 128, 12, 32]);  view_163 = None\0A    permute_40 = torch.ops.aten.permute(view_164, [0, 2, 1, 3]);  view_164 = None\0A    t_62 = torch.ops.aten.t(params_47)\0A    view_165 = torch.ops.aten.view(getitem_60, [128, 384])\0A    addmm_62 = torch.ops.aten.addmm(params_46, view_165, t_62)\0A    view_166 = torch.ops.aten.view(addmm_62, [1, 128, 384]);  addmm_62 = None\0A    view_167 = torch.ops.aten.view(view_166, [1, 128, 12, 32]);  view_166 = None\0A    permute_41 = torch.ops.aten.permute(view_167, [0, 2, 1, 3]);  view_167 = None\0A    view_168 = torch.ops.aten.view(view_161, [1, 128, 12, 32]);  view_161 = None\0A    permute_42 = torch.ops.aten.permute(view_168, [0, 2, 1, 3]);  view_168 = None\0A    transpose_10 = torch.ops.aten.transpose(permute_40, -1, -2);  permute_40 = None\0A    expand_41 = torch.ops.aten.expand(permute_42, [1, 12, 128, 32]);  permute_42 = None\0A    _reshape_alias_40 = torch.ops.aten._reshape_alias(expand_41, [12, 128, 32], [32, 384, 1]);  expand_41 = None\0A    expand_42 = torch.ops.aten.expand(transpose_10, [1, 12, 32, 128]);  transpose_10 = None\0A    _reshape_alias_41 = torch.ops.aten._reshape_alias(expand_42, [12, 32, 128], [32, 1, 384]);  expand_42 = None\0A    bmm_20 = torch.ops.aten.bmm(_reshape_alias_40, _reshape_alias_41)\0A    _unsafe_view_20 = torch.ops.aten._unsafe_view(bmm_20, [1, 12, 128, 128]);  bmm_20 = None\0A    div_10 = torch.ops.aten.div(_unsafe_view_20, 5.656854249492381);  _unsafe_view_20 = None\0A    _tensor_constant0_10 = self._tensor_constant0\0A    add_31 = torch.ops.aten.add(div_10, _tensor_constant0_10);  div_10 = _tensor_constant0_10 = None\0A    _softmax_10 = torch.ops.aten._softmax(add_31, -1, False);  add_31 = None\0A    detach_10 = torch.ops.aten.detach(_softmax_10)\0A    expand_43 = torch.ops.aten.expand(_softmax_10, [1, 12, 128, 128]);  _softmax_10 = None\0A    _reshape_alias_42 = torch.ops.aten._reshape_alias(expand_43, [12, 128, 128], [16384, 128, 1]);  expand_43 = None\0A    expand_44 = torch.ops.aten.expand(permute_41, [1, 12, 128, 32]);  permute_41 = None\0A    _reshape_alias_43 = torch.ops.aten._reshape_alias(expand_44, [12, 128, 32], [32, 384, 1]);  expand_44 = None\0A    bmm_21 = torch.ops.aten.bmm(_reshape_alias_42, _reshape_alias_43)\0A    _unsafe_view_21 = torch.ops.aten._unsafe_view(bmm_21, [1, 12, 128, 32]);  bmm_21 = None\0A    permute_43 = torch.ops.aten.permute(_unsafe_view_21, [0, 2, 1, 3]);  _unsafe_view_21 = None\0A    clone_10 = torch.ops.aten.clone(permute_43, memory_format = 0);  permute_43 = None\0A    view_169 = torch.ops.aten.view(clone_10, [1, 128, 384]);  clone_10 = None\0A    t_63 = torch.ops.aten.t(params_41)\0A    view_170 = torch.ops.aten.view(view_169, [128, 384]);  view_169 = None\0A    addmm_63 = torch.ops.aten.addmm(params_40, view_170, t_63)\0A    view_171 = torch.ops.aten.view(addmm_63, [1, 128, 384]);  addmm_63 = None\0A    add_32 = torch.ops.aten.add(view_171, getitem_60);  view_171 = getitem_60 = None\0A    native_layer_norm_21 = torch.ops.aten.native_layer_norm(add_32, [384], params_39, params_38, 1e-12)\0A    getitem_63 = native_layer_norm_21[0]\0A    getitem_64 = native_layer_norm_21[1]\0A    getitem_65 = native_layer_norm_21[2];  native_layer_norm_21 = None\0A    t_64 = torch.ops.aten.t(params_49)\0A    view_172 = torch.ops.aten.view(getitem_63, [128, 384])\0A    addmm_64 = torch.ops.aten.addmm(params_48, view_172, t_64)\0A    view_173 = torch.ops.aten.view(addmm_64, [1, 128, 1536]);  addmm_64 = None\0A    gelu_10 = torch.ops.aten.gelu(view_173)\0A    t_65 = torch.ops.aten.t(params_53)\0A    view_174 = torch.ops.aten.view(gelu_10, [128, 1536]);  gelu_10 = None\0A    addmm_65 = torch.ops.aten.addmm(params_52, view_174, t_65)\0A    view_175 = torch.ops.aten.view(addmm_65, [1, 128, 384]);  addmm_65 = None\0A    add_33 = torch.ops.aten.add(view_175, getitem_63);  view_175 = getitem_63 = None\0A    native_layer_norm_22 = torch.ops.aten.native_layer_norm(add_33, [384], params_51, params_50, 1e-12)\0A    getitem_66 = native_layer_norm_22[0]\0A    getitem_67 = native_layer_norm_22[1]\0A    getitem_68 = native_layer_norm_22[2];  native_layer_norm_22 = None\0A    t_66 = torch.ops.aten.t(params_61)\0A    view_176 = torch.ops.aten.view(getitem_66, [128, 384])\0A    addmm_66 = torch.ops.aten.addmm(params_60, view_176, t_66)\0A    view_177 = torch.ops.aten.view(addmm_66, [1, 128, 384]);  addmm_66 = None\0A    t_67 = torch.ops.aten.t(params_59)\0A    view_178 = torch.ops.aten.view(getitem_66, [128, 384])\0A    addmm_67 = torch.ops.aten.addmm(params_58, view_178, t_67)\0A    view_179 = torch.ops.aten.view(addmm_67, [1, 128, 384]);  addmm_67 = None\0A    view_180 = torch.ops.aten.view(view_179, [1, 128, 12, 32]);  view_179 = None\0A    permute_44 = torch.ops.aten.permute(view_180, [0, 2, 1, 3]);  view_180 = None\0A    t_68 = torch.ops.aten.t(params_63)\0A    view_181 = torch.ops.aten.view(getitem_66, [128, 384])\0A    addmm_68 = torch.ops.aten.addmm(params_62, view_181, t_68)\0A    view_182 = torch.ops.aten.view(addmm_68, [1, 128, 384]);  addmm_68 = None\0A    view_183 = torch.ops.aten.view(view_182, [1, 128, 12, 32]);  view_182 = None\0A    permute_45 = torch.ops.aten.permute(view_183, [0, 2, 1, 3]);  view_183 = None\0A    view_184 = torch.ops.aten.view(view_177, [1, 128, 12, 32]);  view_177 = None\0A    permute_46 = torch.ops.aten.permute(view_184, [0, 2, 1, 3]);  view_184 = None\0A    transpose_11 = torch.ops.aten.transpose(permute_44, -1, -2);  permute_44 = None\0A    expand_45 = torch.ops.aten.expand(permute_46, [1, 12, 128, 32]);  permute_46 = None\0A    _reshape_alias_44 = torch.ops.aten._reshape_alias(expand_45, [12, 128, 32], [32, 384, 1]);  expand_45 = None\0A    expand_46 = torch.ops.aten.expand(transpose_11, [1, 12, 32, 128]);  transpose_11 = None\0A    _reshape_alias_45 = torch.ops.aten._reshape_alias(expand_46, [12, 32, 128], [32, 1, 384]);  expand_46 = None\0A    bmm_22 = torch.ops.aten.bmm(_reshape_alias_44, _reshape_alias_45)\0A    _unsafe_view_22 = torch.ops.aten._unsafe_view(bmm_22, [1, 12, 128, 128]);  bmm_22 = None\0A    div_11 = torch.ops.aten.div(_unsafe_view_22, 5.656854249492381);  _unsafe_view_22 = None\0A    _tensor_constant0_11 = self._tensor_constant0\0A    add_34 = torch.ops.aten.add(div_11, _tensor_constant0_11);  div_11 = _tensor_constant0_11 = None\0A    _softmax_11 = torch.ops.aten._softmax(add_34, -1, False);  add_34 = None\0A    detach_11 = torch.ops.aten.detach(_softmax_11)\0A    expand_47 = torch.ops.aten.expand(_softmax_11, [1, 12, 128, 128]);  _softmax_11 = None\0A    _reshape_alias_46 = torch.ops.aten._reshape_alias(expand_47, [12, 128, 128], [16384, 128, 1]);  expand_47 = None\0A    expand_48 = torch.ops.aten.expand(permute_45, [1, 12, 128, 32]);  permute_45 = None\0A    _reshape_alias_47 = torch.ops.aten._reshape_alias(expand_48, [12, 128, 32], [32, 384, 1]);  expand_48 = None\0A    bmm_23 = torch.ops.aten.bmm(_reshape_alias_46, _reshape_alias_47)\0A    _unsafe_view_23 = torch.ops.aten._unsafe_view(bmm_23, [1, 12, 128, 32]);  bmm_23 = None\0A    permute_47 = torch.ops.aten.permute(_unsafe_view_23, [0, 2, 1, 3]);  _unsafe_view_23 = None\0A    clone_11 = torch.ops.aten.clone(permute_47, memory_format = 0);  permute_47 = None\0A    view_185 = torch.ops.aten.view(clone_11, [1, 128, 384]);  clone_11 = None\0A    t_69 = torch.ops.aten.t(params_57)\0A    view_186 = torch.ops.aten.view(view_185, [128, 384]);  view_185 = None\0A    addmm_69 = torch.ops.aten.addmm(params_56, view_186, t_69)\0A    view_187 = torch.ops.aten.view(addmm_69, [1, 128, 384]);  addmm_69 = None\0A    add_35 = torch.ops.aten.add(view_187, getitem_66);  view_187 = getitem_66 = None\0A    native_layer_norm_23 = torch.ops.aten.native_layer_norm(add_35, [384], params_55, params_54, 1e-12)\0A    getitem_69 = native_layer_norm_23[0]\0A    getitem_70 = native_layer_norm_23[1]\0A    getitem_71 = native_layer_norm_23[2];  native_layer_norm_23 = None\0A    t_70 = torch.ops.aten.t(params_65)\0A    view_188 = torch.ops.aten.view(getitem_69, [128, 384])\0A    addmm_70 = torch.ops.aten.addmm(params_64, view_188, t_70)\0A    view_189 = torch.ops.aten.view(addmm_70, [1, 128, 1536]);  addmm_70 = None\0A    gelu_11 = torch.ops.aten.gelu(view_189)\0A    t_71 = torch.ops.aten.t(params_69)\0A    view_190 = torch.ops.aten.view(gelu_11, [128, 1536]);  gelu_11 = None\0A    addmm_71 = torch.ops.aten.addmm(params_68, view_190, t_71)\0A    view_191 = torch.ops.aten.view(addmm_71, [1, 128, 384]);  addmm_71 = None\0A    add_36 = torch.ops.aten.add(view_191, getitem_69);  view_191 = getitem_69 = None\0A    native_layer_norm_24 = torch.ops.aten.native_layer_norm(add_36, [384], params_67, params_66, 1e-12)\0A    getitem_72 = native_layer_norm_24[0]\0A    getitem_73 = native_layer_norm_24[1]\0A    getitem_74 = native_layer_norm_24[2];  native_layer_norm_24 = None\0A    slice_5 = torch.ops.aten.slice(getitem_72, 0, 0, 9223372036854775807);  getitem_72 = None\0A    select = torch.ops.aten.select(slice_5, 1, 0);  slice_5 = None\0A    t_72 = torch.ops.aten.t(params_199)\0A    addmm_72 = torch.ops.aten.addmm(params_198, select, t_72)\0A    tanh = torch.ops.aten.tanh(addmm_72);  addmm_72 = None\0A    detach_12 = torch.ops.aten.detach(tanh)\0A    t_73 = torch.ops.aten.t(params_201)\0A    addmm_73 = torch.ops.aten.addmm(params_200, tanh, t_73)\0A    sum_1 = torch.ops.aten.sum(addmm_73);  addmm_73 = None\0A    ones_like = torch.ops.aten.ones_like(sum_1, device = device(type='cpu'), dtype = 6, layout = 0, memory_format = 1, pin_memory = False);  sum_1 = None\0A    expand_49 = torch.ops.aten.expand(ones_like, [1, 2]);  ones_like = None\0A    t_74 = torch.ops.aten.t(t_73);  t_73 = None\0A    mm = torch.ops.aten.mm(expand_49, t_74);  t_74 = None\0A    t_75 = torch.ops.aten.t(expand_49)\0A    mm_1 = torch.ops.aten.mm(t_75, tanh);  t_75 = tanh = None\0A    t_76 = torch.ops.aten.t(mm_1);  mm_1 = None\0A    sum_2 = torch.ops.aten.sum(expand_49, [0], True);  expand_49 = None\0A    view_192 = torch.ops.aten.view(sum_2, [2]);  sum_2 = None\0A    detach_13 = torch.ops.aten.detach(view_192);  view_192 = None\0A    detach_14 = torch.ops.aten.detach(detach_13);  detach_13 = None\0A    t_77 = torch.ops.aten.t(t_76);  t_76 = None\0A    detach_15 = torch.ops.aten.detach(t_77);  t_77 = None\0A    detach_16 = torch.ops.aten.detach(detach_15);  detach_15 = None\0A    detach_17 = torch.ops.aten.detach(detach_12);  detach_12 = None\0A    tanh_backward = torch.ops.aten.tanh_backward(mm, detach_17);  mm = detach_17 = None\0A    t_78 = torch.ops.aten.t(t_72);  t_72 = None\0A    mm_2 = torch.ops.aten.mm(tanh_backward, t_78);  t_78 = None\0A    t_79 = torch.ops.aten.t(tanh_backward)\0A    mm_3 = torch.ops.aten.mm(t_79, select);  t_79 = select = None\0A    t_80 = torch.ops.aten.t(mm_3);  mm_3 = None\0A    sum_3 = torch.ops.aten.sum(tanh_backward, [0], True);  tanh_backward = None\0A    view_193 = torch.ops.aten.view(sum_3, [384]);  sum_3 = None\0A    detach_18 = torch.ops.aten.detach(view_193);  view_193 = None\0A    detach_19 = torch.ops.aten.detach(detach_18);  detach_18 = None\0A    t_81 = torch.ops.aten.t(t_80);  t_80 = None\0A    detach_20 = torch.ops.aten.detach(t_81);  t_81 = None\0A    detach_21 = torch.ops.aten.detach(detach_20);  detach_20 = None\0A    new_zeros = torch.ops.aten.new_zeros(mm_2, [1, 128, 384], device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    select_scatter = torch.ops.aten.select_scatter(new_zeros, mm_2, 1, 0);  new_zeros = mm_2 = None\0A    new_zeros_1 = torch.ops.aten.new_zeros(select_scatter, [1, 128, 384], device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    slice_scatter = torch.ops.aten.slice_scatter(new_zeros_1, select_scatter, 0, 0, 9223372036854775807);  new_zeros_1 = select_scatter = None\0A    to = torch.ops.aten.to(slice_scatter, 6)\0A    to_1 = torch.ops.aten.to(add_36, 6)\0A    to_2 = torch.ops.aten.to(getitem_73, 6)\0A    to_3 = torch.ops.aten.to(getitem_74, 6)\0A    to_4 = torch.ops.aten.to(params_67, 6)\0A    to_5 = torch.ops.aten.to(params_66, 6)\0A    sub = torch.ops.aten.sub(add_36, getitem_73);  add_36 = getitem_73 = None\0A    mul = torch.ops.aten.mul(sub, getitem_74);  sub = None\0A    mul_1 = torch.ops.aten.mul(slice_scatter, params_67)\0A    mul_2 = torch.ops.aten.mul(mul_1, 384)\0A    sum_4 = torch.ops.aten.sum(mul_1, [2], True)\0A    mul_3 = torch.ops.aten.mul(mul_1, mul);  mul_1 = None\0A    sum_5 = torch.ops.aten.sum(mul_3, [2], True);  mul_3 = None\0A    mul_4 = torch.ops.aten.mul(mul, sum_5);  sum_5 = None\0A    sub_1 = torch.ops.aten.sub(mul_2, sum_4);  mul_2 = sum_4 = None\0A    sub_2 = torch.ops.aten.sub(sub_1, mul_4);  sub_1 = mul_4 = None\0A    div_12 = torch.ops.aten.div(getitem_74, 384);  getitem_74 = None\0A    mul_5 = torch.ops.aten.mul(div_12, sub_2);  div_12 = sub_2 = None\0A    mul_6 = torch.ops.aten.mul(slice_scatter, mul);  mul = None\0A    sum_6 = torch.ops.aten.sum(mul_6, [0, 1]);  mul_6 = None\0A    sum_7 = torch.ops.aten.sum(slice_scatter, [0, 1]);  slice_scatter = None\0A    to_6 = torch.ops.aten.to(mul_5, 6)\0A    to_7 = torch.ops.aten.to(sum_6, 6)\0A    to_8 = torch.ops.aten.to(sum_7, 6)\0A    detach_22 = torch.ops.aten.detach(sum_6);  sum_6 = None\0A    detach_23 = torch.ops.aten.detach(detach_22);  detach_22 = None\0A    detach_24 = torch.ops.aten.detach(sum_7);  sum_7 = None\0A    detach_25 = torch.ops.aten.detach(detach_24);  detach_24 = None\0A    _reshape_alias_48 = torch.ops.aten._reshape_alias(mul_5, [128, 384], [384, 1])\0A    t_82 = torch.ops.aten.t(t_71);  t_71 = None\0A    mm_4 = torch.ops.aten.mm(_reshape_alias_48, t_82);  t_82 = None\0A    t_83 = torch.ops.aten.t(_reshape_alias_48)\0A    mm_5 = torch.ops.aten.mm(t_83, view_190);  t_83 = view_190 = None\0A    t_84 = torch.ops.aten.t(mm_5);  mm_5 = None\0A    sum_8 = torch.ops.aten.sum(_reshape_alias_48, [0], True);  _reshape_alias_48 = None\0A    view_194 = torch.ops.aten.view(sum_8, [384]);  sum_8 = None\0A    detach_26 = torch.ops.aten.detach(view_194);  view_194 = None\0A    detach_27 = torch.ops.aten.detach(detach_26);  detach_26 = None\0A    _reshape_alias_49 = torch.ops.aten._reshape_alias(mm_4, [1, 128, 1536], [196608, 1536, 1]);  mm_4 = None\0A    t_85 = torch.ops.aten.t(t_84);  t_84 = None\0A    detach_28 = torch.ops.aten.detach(t_85);  t_85 = None\0A    detach_29 = torch.ops.aten.detach(detach_28);  detach_28 = None\0A    gelu_backward = torch.ops.aten.gelu_backward(_reshape_alias_49, view_189);  _reshape_alias_49 = view_189 = None\0A    _reshape_alias_50 = torch.ops.aten._reshape_alias(gelu_backward, [128, 1536], [1536, 1]);  gelu_backward = None\0A    t_86 = torch.ops.aten.t(t_70);  t_70 = None\0A    mm_6 = torch.ops.aten.mm(_reshape_alias_50, t_86);  t_86 = None\0A    t_87 = torch.ops.aten.t(_reshape_alias_50)\0A    mm_7 = torch.ops.aten.mm(t_87, view_188);  t_87 = view_188 = None\0A    t_88 = torch.ops.aten.t(mm_7);  mm_7 = None\0A    sum_9 = torch.ops.aten.sum(_reshape_alias_50, [0], True);  _reshape_alias_50 = None\0A    view_195 = torch.ops.aten.view(sum_9, [1536]);  sum_9 = None\0A    detach_30 = torch.ops.aten.detach(view_195);  view_195 = None\0A    detach_31 = torch.ops.aten.detach(detach_30);  detach_30 = None\0A    _reshape_alias_51 = torch.ops.aten._reshape_alias(mm_6, [1, 128, 384], [49152, 384, 1]);  mm_6 = None\0A    add_37 = torch.ops.aten.add(mul_5, _reshape_alias_51);  mul_5 = _reshape_alias_51 = None\0A    t_89 = torch.ops.aten.t(t_88);  t_88 = None\0A    detach_32 = torch.ops.aten.detach(t_89);  t_89 = None\0A    detach_33 = torch.ops.aten.detach(detach_32);  detach_32 = None\0A    to_9 = torch.ops.aten.to(add_37, 6)\0A    to_10 = torch.ops.aten.to(add_35, 6)\0A    to_11 = torch.ops.aten.to(getitem_70, 6)\0A    to_12 = torch.ops.aten.to(getitem_71, 6)\0A    to_13 = torch.ops.aten.to(params_55, 6)\0A    to_14 = torch.ops.aten.to(params_54, 6)\0A    sub_3 = torch.ops.aten.sub(add_35, getitem_70);  add_35 = getitem_70 = None\0A    mul_7 = torch.ops.aten.mul(sub_3, getitem_71);  sub_3 = None\0A    mul_8 = torch.ops.aten.mul(add_37, params_55)\0A    mul_9 = torch.ops.aten.mul(mul_8, 384)\0A    sum_10 = torch.ops.aten.sum(mul_8, [2], True)\0A    mul_10 = torch.ops.aten.mul(mul_8, mul_7);  mul_8 = None\0A    sum_11 = torch.ops.aten.sum(mul_10, [2], True);  mul_10 = None\0A    mul_11 = torch.ops.aten.mul(mul_7, sum_11);  sum_11 = None\0A    sub_4 = torch.ops.aten.sub(mul_9, sum_10);  mul_9 = sum_10 = None\0A    sub_5 = torch.ops.aten.sub(sub_4, mul_11);  sub_4 = mul_11 = None\0A    div_13 = torch.ops.aten.div(getitem_71, 384);  getitem_71 = None\0A    mul_12 = torch.ops.aten.mul(div_13, sub_5);  div_13 = sub_5 = None\0A    mul_13 = torch.ops.aten.mul(add_37, mul_7);  mul_7 = None\0A    sum_12 = torch.ops.aten.sum(mul_13, [0, 1]);  mul_13 = None\0A    sum_13 = torch.ops.aten.sum(add_37, [0, 1]);  add_37 = None\0A    to_15 = torch.ops.aten.to(mul_12, 6)\0A    to_16 = torch.ops.aten.to(sum_12, 6)\0A    to_17 = torch.ops.aten.to(sum_13, 6)\0A    detach_34 = torch.ops.aten.detach(sum_12);  sum_12 = None\0A    detach_35 = torch.ops.aten.detach(detach_34);  detach_34 = None\0A    detach_36 = torch.ops.aten.detach(sum_13);  sum_13 = None\0A    detach_37 = torch.ops.aten.detach(detach_36);  detach_36 = None\0A    _reshape_alias_52 = torch.ops.aten._reshape_alias(mul_12, [128, 384], [384, 1])\0A    t_90 = torch.ops.aten.t(t_69);  t_69 = None\0A    mm_8 = torch.ops.aten.mm(_reshape_alias_52, t_90);  t_90 = None\0A    t_91 = torch.ops.aten.t(_reshape_alias_52)\0A    mm_9 = torch.ops.aten.mm(t_91, view_186);  t_91 = view_186 = None\0A    t_92 = torch.ops.aten.t(mm_9);  mm_9 = None\0A    sum_14 = torch.ops.aten.sum(_reshape_alias_52, [0], True);  _reshape_alias_52 = None\0A    view_196 = torch.ops.aten.view(sum_14, [384]);  sum_14 = None\0A    detach_38 = torch.ops.aten.detach(view_196);  view_196 = None\0A    detach_39 = torch.ops.aten.detach(detach_38);  detach_38 = None\0A    _reshape_alias_53 = torch.ops.aten._reshape_alias(mm_8, [1, 128, 384], [49152, 384, 1]);  mm_8 = None\0A    t_93 = torch.ops.aten.t(t_92);  t_92 = None\0A    detach_40 = torch.ops.aten.detach(t_93);  t_93 = None\0A    detach_41 = torch.ops.aten.detach(detach_40);  detach_40 = None\0A    _reshape_alias_54 = torch.ops.aten._reshape_alias(_reshape_alias_53, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_53 = None\0A    permute_48 = torch.ops.aten.permute(_reshape_alias_54, [0, 2, 1, 3]);  _reshape_alias_54 = None\0A    _reshape_alias_55 = torch.ops.aten._reshape_alias(permute_48, [12, 128, 32], [32, 384, 1]);  permute_48 = None\0A    transpose_12 = torch.ops.aten.transpose(_reshape_alias_46, 1, 2);  _reshape_alias_46 = None\0A    bmm_24 = torch.ops.aten.bmm(transpose_12, _reshape_alias_55);  transpose_12 = None\0A    transpose_13 = torch.ops.aten.transpose(_reshape_alias_47, 1, 2);  _reshape_alias_47 = None\0A    bmm_25 = torch.ops.aten.bmm(_reshape_alias_55, transpose_13);  _reshape_alias_55 = transpose_13 = None\0A    _reshape_alias_56 = torch.ops.aten._reshape_alias(bmm_24, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_24 = None\0A    _reshape_alias_57 = torch.ops.aten._reshape_alias(bmm_25, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_25 = None\0A    detach_42 = torch.ops.aten.detach(detach_11);  detach_11 = None\0A    _softmax_backward_data = torch.ops.aten._softmax_backward_data(_reshape_alias_57, detach_42, -1, 6);  _reshape_alias_57 = detach_42 = None\0A    div_14 = torch.ops.aten.div(_softmax_backward_data, 5.656854249492381);  _softmax_backward_data = None\0A    _reshape_alias_58 = torch.ops.aten._reshape_alias(div_14, [12, 128, 128], [16384, 128, 1]);  div_14 = None\0A    transpose_14 = torch.ops.aten.transpose(_reshape_alias_44, 1, 2);  _reshape_alias_44 = None\0A    bmm_26 = torch.ops.aten.bmm(transpose_14, _reshape_alias_58);  transpose_14 = None\0A    transpose_15 = torch.ops.aten.transpose(_reshape_alias_45, 1, 2);  _reshape_alias_45 = None\0A    bmm_27 = torch.ops.aten.bmm(_reshape_alias_58, transpose_15);  _reshape_alias_58 = transpose_15 = None\0A    _reshape_alias_59 = torch.ops.aten._reshape_alias(bmm_26, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_26 = None\0A    _reshape_alias_60 = torch.ops.aten._reshape_alias(bmm_27, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_27 = None\0A    transpose_16 = torch.ops.aten.transpose(_reshape_alias_59, -1, -2);  _reshape_alias_59 = None\0A    permute_49 = torch.ops.aten.permute(_reshape_alias_60, [0, 2, 1, 3]);  _reshape_alias_60 = None\0A    clone_12 = torch.ops.aten.clone(permute_49, memory_format = 0);  permute_49 = None\0A    _unsafe_view_24 = torch.ops.aten._unsafe_view(clone_12, [1, 128, 384]);  clone_12 = None\0A    permute_50 = torch.ops.aten.permute(_reshape_alias_56, [0, 2, 1, 3]);  _reshape_alias_56 = None\0A    clone_13 = torch.ops.aten.clone(permute_50, memory_format = 0);  permute_50 = None\0A    _unsafe_view_25 = torch.ops.aten._unsafe_view(clone_13, [1, 128, 384]);  clone_13 = None\0A    _reshape_alias_61 = torch.ops.aten._reshape_alias(_unsafe_view_25, [128, 384], [384, 1]);  _unsafe_view_25 = None\0A    t_94 = torch.ops.aten.t(t_68);  t_68 = None\0A    mm_10 = torch.ops.aten.mm(_reshape_alias_61, t_94);  t_94 = None\0A    t_95 = torch.ops.aten.t(_reshape_alias_61)\0A    mm_11 = torch.ops.aten.mm(t_95, view_181);  t_95 = view_181 = None\0A    t_96 = torch.ops.aten.t(mm_11);  mm_11 = None\0A    sum_15 = torch.ops.aten.sum(_reshape_alias_61, [0], True);  _reshape_alias_61 = None\0A    view_197 = torch.ops.aten.view(sum_15, [384]);  sum_15 = None\0A    detach_43 = torch.ops.aten.detach(view_197);  view_197 = None\0A    detach_44 = torch.ops.aten.detach(detach_43);  detach_43 = None\0A    _reshape_alias_62 = torch.ops.aten._reshape_alias(mm_10, [1, 128, 384], [49152, 384, 1]);  mm_10 = None\0A    add_38 = torch.ops.aten.add(mul_12, _reshape_alias_62);  mul_12 = _reshape_alias_62 = None\0A    t_97 = torch.ops.aten.t(t_96);  t_96 = None\0A    detach_45 = torch.ops.aten.detach(t_97);  t_97 = None\0A    detach_46 = torch.ops.aten.detach(detach_45);  detach_45 = None\0A    permute_51 = torch.ops.aten.permute(transpose_16, [0, 2, 1, 3]);  transpose_16 = None\0A    _reshape_alias_63 = torch.ops.aten._reshape_alias(permute_51, [1, 128, 384], [128, 1, 128]);  permute_51 = None\0A    _reshape_alias_64 = torch.ops.aten._reshape_alias(_reshape_alias_63, [128, 384], [1, 128]);  _reshape_alias_63 = None\0A    t_98 = torch.ops.aten.t(t_67);  t_67 = None\0A    mm_12 = torch.ops.aten.mm(_reshape_alias_64, t_98);  t_98 = None\0A    t_99 = torch.ops.aten.t(_reshape_alias_64)\0A    mm_13 = torch.ops.aten.mm(t_99, view_178);  t_99 = view_178 = None\0A    t_100 = torch.ops.aten.t(mm_13);  mm_13 = None\0A    sum_16 = torch.ops.aten.sum(_reshape_alias_64, [0], True);  _reshape_alias_64 = None\0A    view_198 = torch.ops.aten.view(sum_16, [384]);  sum_16 = None\0A    detach_47 = torch.ops.aten.detach(view_198);  view_198 = None\0A    detach_48 = torch.ops.aten.detach(detach_47);  detach_47 = None\0A    _reshape_alias_65 = torch.ops.aten._reshape_alias(mm_12, [1, 128, 384], [49152, 384, 1]);  mm_12 = None\0A    add_39 = torch.ops.aten.add(add_38, _reshape_alias_65);  add_38 = _reshape_alias_65 = None\0A    t_101 = torch.ops.aten.t(t_100);  t_100 = None\0A    detach_49 = torch.ops.aten.detach(t_101);  t_101 = None\0A    detach_50 = torch.ops.aten.detach(detach_49);  detach_49 = None\0A    _reshape_alias_66 = torch.ops.aten._reshape_alias(_unsafe_view_24, [128, 384], [384, 1]);  _unsafe_view_24 = None\0A    t_102 = torch.ops.aten.t(t_66);  t_66 = None\0A    mm_14 = torch.ops.aten.mm(_reshape_alias_66, t_102);  t_102 = None\0A    t_103 = torch.ops.aten.t(_reshape_alias_66)\0A    mm_15 = torch.ops.aten.mm(t_103, view_176);  t_103 = view_176 = None\0A    t_104 = torch.ops.aten.t(mm_15);  mm_15 = None\0A    sum_17 = torch.ops.aten.sum(_reshape_alias_66, [0], True);  _reshape_alias_66 = None\0A    view_199 = torch.ops.aten.view(sum_17, [384]);  sum_17 = None\0A    detach_51 = torch.ops.aten.detach(view_199);  view_199 = None\0A    detach_52 = torch.ops.aten.detach(detach_51);  detach_51 = None\0A    _reshape_alias_67 = torch.ops.aten._reshape_alias(mm_14, [1, 128, 384], [49152, 384, 1]);  mm_14 = None\0A    add_40 = torch.ops.aten.add(add_39, _reshape_alias_67);  add_39 = _reshape_alias_67 = None\0A    t_105 = torch.ops.aten.t(t_104);  t_104 = None\0A    detach_53 = torch.ops.aten.detach(t_105);  t_105 = None\0A    detach_54 = torch.ops.aten.detach(detach_53);  detach_53 = None\0A    to_18 = torch.ops.aten.to(add_40, 6)\0A    to_19 = torch.ops.aten.to(add_33, 6)\0A    to_20 = torch.ops.aten.to(getitem_67, 6)\0A    to_21 = torch.ops.aten.to(getitem_68, 6)\0A    to_22 = torch.ops.aten.to(params_51, 6)\0A    to_23 = torch.ops.aten.to(params_50, 6)\0A    sub_6 = torch.ops.aten.sub(add_33, getitem_67);  add_33 = getitem_67 = None\0A    mul_14 = torch.ops.aten.mul(sub_6, getitem_68);  sub_6 = None\0A    mul_15 = torch.ops.aten.mul(add_40, params_51)\0A    mul_16 = torch.ops.aten.mul(mul_15, 384)\0A    sum_18 = torch.ops.aten.sum(mul_15, [2], True)\0A    mul_17 = torch.ops.aten.mul(mul_15, mul_14);  mul_15 = None\0A    sum_19 = torch.ops.aten.sum(mul_17, [2], True);  mul_17 = None\0A    mul_18 = torch.ops.aten.mul(mul_14, sum_19);  sum_19 = None\0A    sub_7 = torch.ops.aten.sub(mul_16, sum_18);  mul_16 = sum_18 = None\0A    sub_8 = torch.ops.aten.sub(sub_7, mul_18);  sub_7 = mul_18 = None\0A    div_15 = torch.ops.aten.div(getitem_68, 384);  getitem_68 = None\0A    mul_19 = torch.ops.aten.mul(div_15, sub_8);  div_15 = sub_8 = None\0A    mul_20 = torch.ops.aten.mul(add_40, mul_14);  mul_14 = None\0A    sum_20 = torch.ops.aten.sum(mul_20, [0, 1]);  mul_20 = None\0A    sum_21 = torch.ops.aten.sum(add_40, [0, 1]);  add_40 = None\0A    to_24 = torch.ops.aten.to(mul_19, 6)\0A    to_25 = torch.ops.aten.to(sum_20, 6)\0A    to_26 = torch.ops.aten.to(sum_21, 6)\0A    detach_55 = torch.ops.aten.detach(sum_20);  sum_20 = None\0A    detach_56 = torch.ops.aten.detach(detach_55);  detach_55 = None\0A    detach_57 = torch.ops.aten.detach(sum_21);  sum_21 = None\0A    detach_58 = torch.ops.aten.detach(detach_57);  detach_57 = None\0A    _reshape_alias_68 = torch.ops.aten._reshape_alias(mul_19, [128, 384], [384, 1])\0A    t_106 = torch.ops.aten.t(t_65);  t_65 = None\0A    mm_16 = torch.ops.aten.mm(_reshape_alias_68, t_106);  t_106 = None\0A    t_107 = torch.ops.aten.t(_reshape_alias_68)\0A    mm_17 = torch.ops.aten.mm(t_107, view_174);  t_107 = view_174 = None\0A    t_108 = torch.ops.aten.t(mm_17);  mm_17 = None\0A    sum_22 = torch.ops.aten.sum(_reshape_alias_68, [0], True);  _reshape_alias_68 = None\0A    view_200 = torch.ops.aten.view(sum_22, [384]);  sum_22 = None\0A    detach_59 = torch.ops.aten.detach(view_200);  view_200 = None\0A    detach_60 = torch.ops.aten.detach(detach_59);  detach_59 = None\0A    _reshape_alias_69 = torch.ops.aten._reshape_alias(mm_16, [1, 128, 1536], [196608, 1536, 1]);  mm_16 = None\0A    t_109 = torch.ops.aten.t(t_108);  t_108 = None\0A    detach_61 = torch.ops.aten.detach(t_109);  t_109 = None\0A    detach_62 = torch.ops.aten.detach(detach_61);  detach_61 = None\0A    gelu_backward_1 = torch.ops.aten.gelu_backward(_reshape_alias_69, view_173);  _reshape_alias_69 = view_173 = None\0A    _reshape_alias_70 = torch.ops.aten._reshape_alias(gelu_backward_1, [128, 1536], [1536, 1]);  gelu_backward_1 = None\0A    t_110 = torch.ops.aten.t(t_64);  t_64 = None\0A    mm_18 = torch.ops.aten.mm(_reshape_alias_70, t_110);  t_110 = None\0A    t_111 = torch.ops.aten.t(_reshape_alias_70)\0A    mm_19 = torch.ops.aten.mm(t_111, view_172);  t_111 = view_172 = None\0A    t_112 = torch.ops.aten.t(mm_19);  mm_19 = None\0A    sum_23 = torch.ops.aten.sum(_reshape_alias_70, [0], True);  _reshape_alias_70 = None\0A    view_201 = torch.ops.aten.view(sum_23, [1536]);  sum_23 = None\0A    detach_63 = torch.ops.aten.detach(view_201);  view_201 = None\0A    detach_64 = torch.ops.aten.detach(detach_63);  detach_63 = None\0A    _reshape_alias_71 = torch.ops.aten._reshape_alias(mm_18, [1, 128, 384], [49152, 384, 1]);  mm_18 = None\0A    add_41 = torch.ops.aten.add(mul_19, _reshape_alias_71);  mul_19 = _reshape_alias_71 = None\0A    t_113 = torch.ops.aten.t(t_112);  t_112 = None\0A    detach_65 = torch.ops.aten.detach(t_113);  t_113 = None\0A    detach_66 = torch.ops.aten.detach(detach_65);  detach_65 = None\0A    to_27 = torch.ops.aten.to(add_41, 6)\0A    to_28 = torch.ops.aten.to(add_32, 6)\0A    to_29 = torch.ops.aten.to(getitem_64, 6)\0A    to_30 = torch.ops.aten.to(getitem_65, 6)\0A    to_31 = torch.ops.aten.to(params_39, 6)\0A    to_32 = torch.ops.aten.to(params_38, 6)\0A    sub_9 = torch.ops.aten.sub(add_32, getitem_64);  add_32 = getitem_64 = None\0A    mul_21 = torch.ops.aten.mul(sub_9, getitem_65);  sub_9 = None\0A    mul_22 = torch.ops.aten.mul(add_41, params_39)\0A    mul_23 = torch.ops.aten.mul(mul_22, 384)\0A    sum_24 = torch.ops.aten.sum(mul_22, [2], True)\0A    mul_24 = torch.ops.aten.mul(mul_22, mul_21);  mul_22 = None\0A    sum_25 = torch.ops.aten.sum(mul_24, [2], True);  mul_24 = None\0A    mul_25 = torch.ops.aten.mul(mul_21, sum_25);  sum_25 = None\0A    sub_10 = torch.ops.aten.sub(mul_23, sum_24);  mul_23 = sum_24 = None\0A    sub_11 = torch.ops.aten.sub(sub_10, mul_25);  sub_10 = mul_25 = None\0A    div_16 = torch.ops.aten.div(getitem_65, 384);  getitem_65 = None\0A    mul_26 = torch.ops.aten.mul(div_16, sub_11);  div_16 = sub_11 = None\0A    mul_27 = torch.ops.aten.mul(add_41, mul_21);  mul_21 = None\0A    sum_26 = torch.ops.aten.sum(mul_27, [0, 1]);  mul_27 = None\0A    sum_27 = torch.ops.aten.sum(add_41, [0, 1]);  add_41 = None\0A    to_33 = torch.ops.aten.to(mul_26, 6)\0A    to_34 = torch.ops.aten.to(sum_26, 6)\0A    to_35 = torch.ops.aten.to(sum_27, 6)\0A    detach_67 = torch.ops.aten.detach(sum_26);  sum_26 = None\0A    detach_68 = torch.ops.aten.detach(detach_67);  detach_67 = None\0A    detach_69 = torch.ops.aten.detach(sum_27);  sum_27 = None\0A    detach_70 = torch.ops.aten.detach(detach_69);  detach_69 = None\0A    _reshape_alias_72 = torch.ops.aten._reshape_alias(mul_26, [128, 384], [384, 1])\0A    t_114 = torch.ops.aten.t(t_63);  t_63 = None\0A    mm_20 = torch.ops.aten.mm(_reshape_alias_72, t_114);  t_114 = None\0A    t_115 = torch.ops.aten.t(_reshape_alias_72)\0A    mm_21 = torch.ops.aten.mm(t_115, view_170);  t_115 = view_170 = None\0A    t_116 = torch.ops.aten.t(mm_21);  mm_21 = None\0A    sum_28 = torch.ops.aten.sum(_reshape_alias_72, [0], True);  _reshape_alias_72 = None\0A    view_202 = torch.ops.aten.view(sum_28, [384]);  sum_28 = None\0A    detach_71 = torch.ops.aten.detach(view_202);  view_202 = None\0A    detach_72 = torch.ops.aten.detach(detach_71);  detach_71 = None\0A    _reshape_alias_73 = torch.ops.aten._reshape_alias(mm_20, [1, 128, 384], [49152, 384, 1]);  mm_20 = None\0A    t_117 = torch.ops.aten.t(t_116);  t_116 = None\0A    detach_73 = torch.ops.aten.detach(t_117);  t_117 = None\0A    detach_74 = torch.ops.aten.detach(detach_73);  detach_73 = None\0A    _reshape_alias_74 = torch.ops.aten._reshape_alias(_reshape_alias_73, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_73 = None\0A    permute_52 = torch.ops.aten.permute(_reshape_alias_74, [0, 2, 1, 3]);  _reshape_alias_74 = None\0A    _reshape_alias_75 = torch.ops.aten._reshape_alias(permute_52, [12, 128, 32], [32, 384, 1]);  permute_52 = None\0A    transpose_17 = torch.ops.aten.transpose(_reshape_alias_42, 1, 2);  _reshape_alias_42 = None\0A    bmm_28 = torch.ops.aten.bmm(transpose_17, _reshape_alias_75);  transpose_17 = None\0A    transpose_18 = torch.ops.aten.transpose(_reshape_alias_43, 1, 2);  _reshape_alias_43 = None\0A    bmm_29 = torch.ops.aten.bmm(_reshape_alias_75, transpose_18);  _reshape_alias_75 = transpose_18 = None\0A    _reshape_alias_76 = torch.ops.aten._reshape_alias(bmm_28, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_28 = None\0A    _reshape_alias_77 = torch.ops.aten._reshape_alias(bmm_29, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_29 = None\0A    detach_75 = torch.ops.aten.detach(detach_10);  detach_10 = None\0A    _softmax_backward_data_1 = torch.ops.aten._softmax_backward_data(_reshape_alias_77, detach_75, -1, 6);  _reshape_alias_77 = detach_75 = None\0A    div_17 = torch.ops.aten.div(_softmax_backward_data_1, 5.656854249492381);  _softmax_backward_data_1 = None\0A    _reshape_alias_78 = torch.ops.aten._reshape_alias(div_17, [12, 128, 128], [16384, 128, 1]);  div_17 = None\0A    transpose_19 = torch.ops.aten.transpose(_reshape_alias_40, 1, 2);  _reshape_alias_40 = None\0A    bmm_30 = torch.ops.aten.bmm(transpose_19, _reshape_alias_78);  transpose_19 = None\0A    transpose_20 = torch.ops.aten.transpose(_reshape_alias_41, 1, 2);  _reshape_alias_41 = None\0A    bmm_31 = torch.ops.aten.bmm(_reshape_alias_78, transpose_20);  _reshape_alias_78 = transpose_20 = None\0A    _reshape_alias_79 = torch.ops.aten._reshape_alias(bmm_30, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_30 = None\0A    _reshape_alias_80 = torch.ops.aten._reshape_alias(bmm_31, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_31 = None\0A    transpose_21 = torch.ops.aten.transpose(_reshape_alias_79, -1, -2);  _reshape_alias_79 = None\0A    permute_53 = torch.ops.aten.permute(_reshape_alias_80, [0, 2, 1, 3]);  _reshape_alias_80 = None\0A    clone_14 = torch.ops.aten.clone(permute_53, memory_format = 0);  permute_53 = None\0A    _unsafe_view_26 = torch.ops.aten._unsafe_view(clone_14, [1, 128, 384]);  clone_14 = None\0A    permute_54 = torch.ops.aten.permute(_reshape_alias_76, [0, 2, 1, 3]);  _reshape_alias_76 = None\0A    clone_15 = torch.ops.aten.clone(permute_54, memory_format = 0);  permute_54 = None\0A    _unsafe_view_27 = torch.ops.aten._unsafe_view(clone_15, [1, 128, 384]);  clone_15 = None\0A    _reshape_alias_81 = torch.ops.aten._reshape_alias(_unsafe_view_27, [128, 384], [384, 1]);  _unsafe_view_27 = None\0A    t_118 = torch.ops.aten.t(t_62);  t_62 = None\0A    mm_22 = torch.ops.aten.mm(_reshape_alias_81, t_118);  t_118 = None\0A    t_119 = torch.ops.aten.t(_reshape_alias_81)\0A    mm_23 = torch.ops.aten.mm(t_119, view_165);  t_119 = view_165 = None\0A    t_120 = torch.ops.aten.t(mm_23);  mm_23 = None\0A    sum_29 = torch.ops.aten.sum(_reshape_alias_81, [0], True);  _reshape_alias_81 = None\0A    view_203 = torch.ops.aten.view(sum_29, [384]);  sum_29 = None\0A    detach_76 = torch.ops.aten.detach(view_203);  view_203 = None\0A    detach_77 = torch.ops.aten.detach(detach_76);  detach_76 = None\0A    _reshape_alias_82 = torch.ops.aten._reshape_alias(mm_22, [1, 128, 384], [49152, 384, 1]);  mm_22 = None\0A    add_42 = torch.ops.aten.add(mul_26, _reshape_alias_82);  mul_26 = _reshape_alias_82 = None\0A    t_121 = torch.ops.aten.t(t_120);  t_120 = None\0A    detach_78 = torch.ops.aten.detach(t_121);  t_121 = None\0A    detach_79 = torch.ops.aten.detach(detach_78);  detach_78 = None\0A    permute_55 = torch.ops.aten.permute(transpose_21, [0, 2, 1, 3]);  transpose_21 = None\0A    _reshape_alias_83 = torch.ops.aten._reshape_alias(permute_55, [1, 128, 384], [128, 1, 128]);  permute_55 = None\0A    _reshape_alias_84 = torch.ops.aten._reshape_alias(_reshape_alias_83, [128, 384], [1, 128]);  _reshape_alias_83 = None\0A    t_122 = torch.ops.aten.t(t_61);  t_61 = None\0A    mm_24 = torch.ops.aten.mm(_reshape_alias_84, t_122);  t_122 = None\0A    t_123 = torch.ops.aten.t(_reshape_alias_84)\0A    mm_25 = torch.ops.aten.mm(t_123, view_162);  t_123 = view_162 = None\0A    t_124 = torch.ops.aten.t(mm_25);  mm_25 = None\0A    sum_30 = torch.ops.aten.sum(_reshape_alias_84, [0], True);  _reshape_alias_84 = None\0A    view_204 = torch.ops.aten.view(sum_30, [384]);  sum_30 = None\0A    detach_80 = torch.ops.aten.detach(view_204);  view_204 = None\0A    detach_81 = torch.ops.aten.detach(detach_80);  detach_80 = None\0A    _reshape_alias_85 = torch.ops.aten._reshape_alias(mm_24, [1, 128, 384], [49152, 384, 1]);  mm_24 = None\0A    add_43 = torch.ops.aten.add(add_42, _reshape_alias_85);  add_42 = _reshape_alias_85 = None\0A    t_125 = torch.ops.aten.t(t_124);  t_124 = None\0A    detach_82 = torch.ops.aten.detach(t_125);  t_125 = None\0A    detach_83 = torch.ops.aten.detach(detach_82);  detach_82 = None\0A    _reshape_alias_86 = torch.ops.aten._reshape_alias(_unsafe_view_26, [128, 384], [384, 1]);  _unsafe_view_26 = None\0A    t_126 = torch.ops.aten.t(t_60);  t_60 = None\0A    mm_26 = torch.ops.aten.mm(_reshape_alias_86, t_126);  t_126 = None\0A    t_127 = torch.ops.aten.t(_reshape_alias_86)\0A    mm_27 = torch.ops.aten.mm(t_127, view_160);  t_127 = view_160 = None\0A    t_128 = torch.ops.aten.t(mm_27);  mm_27 = None\0A    sum_31 = torch.ops.aten.sum(_reshape_alias_86, [0], True);  _reshape_alias_86 = None\0A    view_205 = torch.ops.aten.view(sum_31, [384]);  sum_31 = None\0A    detach_84 = torch.ops.aten.detach(view_205);  view_205 = None\0A    detach_85 = torch.ops.aten.detach(detach_84);  detach_84 = None\0A    _reshape_alias_87 = torch.ops.aten._reshape_alias(mm_26, [1, 128, 384], [49152, 384, 1]);  mm_26 = None\0A    add_44 = torch.ops.aten.add(add_43, _reshape_alias_87);  add_43 = _reshape_alias_87 = None\0A    t_129 = torch.ops.aten.t(t_128);  t_128 = None\0A    detach_86 = torch.ops.aten.detach(t_129);  t_129 = None\0A    detach_87 = torch.ops.aten.detach(detach_86);  detach_86 = None\0A    to_36 = torch.ops.aten.to(add_44, 6)\0A    to_37 = torch.ops.aten.to(add_30, 6)\0A    to_38 = torch.ops.aten.to(getitem_61, 6)\0A    to_39 = torch.ops.aten.to(getitem_62, 6)\0A    to_40 = torch.ops.aten.to(params_195, 6)\0A    to_41 = torch.ops.aten.to(params_194, 6)\0A    sub_12 = torch.ops.aten.sub(add_30, getitem_61);  add_30 = getitem_61 = None\0A    mul_28 = torch.ops.aten.mul(sub_12, getitem_62);  sub_12 = None\0A    mul_29 = torch.ops.aten.mul(add_44, params_195)\0A    mul_30 = torch.ops.aten.mul(mul_29, 384)\0A    sum_32 = torch.ops.aten.sum(mul_29, [2], True)\0A    mul_31 = torch.ops.aten.mul(mul_29, mul_28);  mul_29 = None\0A    sum_33 = torch.ops.aten.sum(mul_31, [2], True);  mul_31 = None\0A    mul_32 = torch.ops.aten.mul(mul_28, sum_33);  sum_33 = None\0A    sub_13 = torch.ops.aten.sub(mul_30, sum_32);  mul_30 = sum_32 = None\0A    sub_14 = torch.ops.aten.sub(sub_13, mul_32);  sub_13 = mul_32 = None\0A    div_18 = torch.ops.aten.div(getitem_62, 384);  getitem_62 = None\0A    mul_33 = torch.ops.aten.mul(div_18, sub_14);  div_18 = sub_14 = None\0A    mul_34 = torch.ops.aten.mul(add_44, mul_28);  mul_28 = None\0A    sum_34 = torch.ops.aten.sum(mul_34, [0, 1]);  mul_34 = None\0A    sum_35 = torch.ops.aten.sum(add_44, [0, 1]);  add_44 = None\0A    to_42 = torch.ops.aten.to(mul_33, 6)\0A    to_43 = torch.ops.aten.to(sum_34, 6)\0A    to_44 = torch.ops.aten.to(sum_35, 6)\0A    detach_88 = torch.ops.aten.detach(sum_34);  sum_34 = None\0A    detach_89 = torch.ops.aten.detach(detach_88);  detach_88 = None\0A    detach_90 = torch.ops.aten.detach(sum_35);  sum_35 = None\0A    detach_91 = torch.ops.aten.detach(detach_90);  detach_90 = None\0A    _reshape_alias_88 = torch.ops.aten._reshape_alias(mul_33, [128, 384], [384, 1])\0A    t_130 = torch.ops.aten.t(t_59);  t_59 = None\0A    mm_28 = torch.ops.aten.mm(_reshape_alias_88, t_130);  t_130 = None\0A    t_131 = torch.ops.aten.t(_reshape_alias_88)\0A    mm_29 = torch.ops.aten.mm(t_131, view_158);  t_131 = view_158 = None\0A    t_132 = torch.ops.aten.t(mm_29);  mm_29 = None\0A    sum_36 = torch.ops.aten.sum(_reshape_alias_88, [0], True);  _reshape_alias_88 = None\0A    view_206 = torch.ops.aten.view(sum_36, [384]);  sum_36 = None\0A    detach_92 = torch.ops.aten.detach(view_206);  view_206 = None\0A    detach_93 = torch.ops.aten.detach(detach_92);  detach_92 = None\0A    _reshape_alias_89 = torch.ops.aten._reshape_alias(mm_28, [1, 128, 1536], [196608, 1536, 1]);  mm_28 = None\0A    t_133 = torch.ops.aten.t(t_132);  t_132 = None\0A    detach_94 = torch.ops.aten.detach(t_133);  t_133 = None\0A    detach_95 = torch.ops.aten.detach(detach_94);  detach_94 = None\0A    gelu_backward_2 = torch.ops.aten.gelu_backward(_reshape_alias_89, view_157);  _reshape_alias_89 = view_157 = None\0A    _reshape_alias_90 = torch.ops.aten._reshape_alias(gelu_backward_2, [128, 1536], [1536, 1]);  gelu_backward_2 = None\0A    t_134 = torch.ops.aten.t(t_58);  t_58 = None\0A    mm_30 = torch.ops.aten.mm(_reshape_alias_90, t_134);  t_134 = None\0A    t_135 = torch.ops.aten.t(_reshape_alias_90)\0A    mm_31 = torch.ops.aten.mm(t_135, view_156);  t_135 = view_156 = None\0A    t_136 = torch.ops.aten.t(mm_31);  mm_31 = None\0A    sum_37 = torch.ops.aten.sum(_reshape_alias_90, [0], True);  _reshape_alias_90 = None\0A    view_207 = torch.ops.aten.view(sum_37, [1536]);  sum_37 = None\0A    detach_96 = torch.ops.aten.detach(view_207);  view_207 = None\0A    detach_97 = torch.ops.aten.detach(detach_96);  detach_96 = None\0A    _reshape_alias_91 = torch.ops.aten._reshape_alias(mm_30, [1, 128, 384], [49152, 384, 1]);  mm_30 = None\0A    add_45 = torch.ops.aten.add(mul_33, _reshape_alias_91);  mul_33 = _reshape_alias_91 = None\0A    t_137 = torch.ops.aten.t(t_136);  t_136 = None\0A    detach_98 = torch.ops.aten.detach(t_137);  t_137 = None\0A    detach_99 = torch.ops.aten.detach(detach_98);  detach_98 = None\0A    to_45 = torch.ops.aten.to(add_45, 6)\0A    to_46 = torch.ops.aten.to(add_29, 6)\0A    to_47 = torch.ops.aten.to(getitem_58, 6)\0A    to_48 = torch.ops.aten.to(getitem_59, 6)\0A    to_49 = torch.ops.aten.to(params_183, 6)\0A    to_50 = torch.ops.aten.to(params_182, 6)\0A    sub_15 = torch.ops.aten.sub(add_29, getitem_58);  add_29 = getitem_58 = None\0A    mul_35 = torch.ops.aten.mul(sub_15, getitem_59);  sub_15 = None\0A    mul_36 = torch.ops.aten.mul(add_45, params_183)\0A    mul_37 = torch.ops.aten.mul(mul_36, 384)\0A    sum_38 = torch.ops.aten.sum(mul_36, [2], True)\0A    mul_38 = torch.ops.aten.mul(mul_36, mul_35);  mul_36 = None\0A    sum_39 = torch.ops.aten.sum(mul_38, [2], True);  mul_38 = None\0A    mul_39 = torch.ops.aten.mul(mul_35, sum_39);  sum_39 = None\0A    sub_16 = torch.ops.aten.sub(mul_37, sum_38);  mul_37 = sum_38 = None\0A    sub_17 = torch.ops.aten.sub(sub_16, mul_39);  sub_16 = mul_39 = None\0A    div_19 = torch.ops.aten.div(getitem_59, 384);  getitem_59 = None\0A    mul_40 = torch.ops.aten.mul(div_19, sub_17);  div_19 = sub_17 = None\0A    mul_41 = torch.ops.aten.mul(add_45, mul_35);  mul_35 = None\0A    sum_40 = torch.ops.aten.sum(mul_41, [0, 1]);  mul_41 = None\0A    sum_41 = torch.ops.aten.sum(add_45, [0, 1]);  add_45 = None\0A    to_51 = torch.ops.aten.to(mul_40, 6)\0A    to_52 = torch.ops.aten.to(sum_40, 6)\0A    to_53 = torch.ops.aten.to(sum_41, 6)\0A    detach_100 = torch.ops.aten.detach(sum_40);  sum_40 = None\0A    detach_101 = torch.ops.aten.detach(detach_100);  detach_100 = None\0A    detach_102 = torch.ops.aten.detach(sum_41);  sum_41 = None\0A    detach_103 = torch.ops.aten.detach(detach_102);  detach_102 = None\0A    _reshape_alias_92 = torch.ops.aten._reshape_alias(mul_40, [128, 384], [384, 1])\0A    t_138 = torch.ops.aten.t(t_57);  t_57 = None\0A    mm_32 = torch.ops.aten.mm(_reshape_alias_92, t_138);  t_138 = None\0A    t_139 = torch.ops.aten.t(_reshape_alias_92)\0A    mm_33 = torch.ops.aten.mm(t_139, view_154);  t_139 = view_154 = None\0A    t_140 = torch.ops.aten.t(mm_33);  mm_33 = None\0A    sum_42 = torch.ops.aten.sum(_reshape_alias_92, [0], True);  _reshape_alias_92 = None\0A    view_208 = torch.ops.aten.view(sum_42, [384]);  sum_42 = None\0A    detach_104 = torch.ops.aten.detach(view_208);  view_208 = None\0A    detach_105 = torch.ops.aten.detach(detach_104);  detach_104 = None\0A    _reshape_alias_93 = torch.ops.aten._reshape_alias(mm_32, [1, 128, 384], [49152, 384, 1]);  mm_32 = None\0A    t_141 = torch.ops.aten.t(t_140);  t_140 = None\0A    detach_106 = torch.ops.aten.detach(t_141);  t_141 = None\0A    detach_107 = torch.ops.aten.detach(detach_106);  detach_106 = None\0A    _reshape_alias_94 = torch.ops.aten._reshape_alias(_reshape_alias_93, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_93 = None\0A    permute_56 = torch.ops.aten.permute(_reshape_alias_94, [0, 2, 1, 3]);  _reshape_alias_94 = None\0A    _reshape_alias_95 = torch.ops.aten._reshape_alias(permute_56, [12, 128, 32], [32, 384, 1]);  permute_56 = None\0A    transpose_22 = torch.ops.aten.transpose(_reshape_alias_38, 1, 2);  _reshape_alias_38 = None\0A    bmm_32 = torch.ops.aten.bmm(transpose_22, _reshape_alias_95);  transpose_22 = None\0A    transpose_23 = torch.ops.aten.transpose(_reshape_alias_39, 1, 2);  _reshape_alias_39 = None\0A    bmm_33 = torch.ops.aten.bmm(_reshape_alias_95, transpose_23);  _reshape_alias_95 = transpose_23 = None\0A    _reshape_alias_96 = torch.ops.aten._reshape_alias(bmm_32, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_32 = None\0A    _reshape_alias_97 = torch.ops.aten._reshape_alias(bmm_33, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_33 = None\0A    detach_108 = torch.ops.aten.detach(detach_9);  detach_9 = None\0A    _softmax_backward_data_2 = torch.ops.aten._softmax_backward_data(_reshape_alias_97, detach_108, -1, 6);  _reshape_alias_97 = detach_108 = None\0A    div_20 = torch.ops.aten.div(_softmax_backward_data_2, 5.656854249492381);  _softmax_backward_data_2 = None\0A    _reshape_alias_98 = torch.ops.aten._reshape_alias(div_20, [12, 128, 128], [16384, 128, 1]);  div_20 = None\0A    transpose_24 = torch.ops.aten.transpose(_reshape_alias_36, 1, 2);  _reshape_alias_36 = None\0A    bmm_34 = torch.ops.aten.bmm(transpose_24, _reshape_alias_98);  transpose_24 = None\0A    transpose_25 = torch.ops.aten.transpose(_reshape_alias_37, 1, 2);  _reshape_alias_37 = None\0A    bmm_35 = torch.ops.aten.bmm(_reshape_alias_98, transpose_25);  _reshape_alias_98 = transpose_25 = None\0A    _reshape_alias_99 = torch.ops.aten._reshape_alias(bmm_34, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_34 = None\0A    _reshape_alias_100 = torch.ops.aten._reshape_alias(bmm_35, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_35 = None\0A    transpose_26 = torch.ops.aten.transpose(_reshape_alias_99, -1, -2);  _reshape_alias_99 = None\0A    permute_57 = torch.ops.aten.permute(_reshape_alias_100, [0, 2, 1, 3]);  _reshape_alias_100 = None\0A    clone_16 = torch.ops.aten.clone(permute_57, memory_format = 0);  permute_57 = None\0A    _unsafe_view_28 = torch.ops.aten._unsafe_view(clone_16, [1, 128, 384]);  clone_16 = None\0A    permute_58 = torch.ops.aten.permute(_reshape_alias_96, [0, 2, 1, 3]);  _reshape_alias_96 = None\0A    clone_17 = torch.ops.aten.clone(permute_58, memory_format = 0);  permute_58 = None\0A    _unsafe_view_29 = torch.ops.aten._unsafe_view(clone_17, [1, 128, 384]);  clone_17 = None\0A    _reshape_alias_101 = torch.ops.aten._reshape_alias(_unsafe_view_29, [128, 384], [384, 1]);  _unsafe_view_29 = None\0A    t_142 = torch.ops.aten.t(t_56);  t_56 = None\0A    mm_34 = torch.ops.aten.mm(_reshape_alias_101, t_142);  t_142 = None\0A    t_143 = torch.ops.aten.t(_reshape_alias_101)\0A    mm_35 = torch.ops.aten.mm(t_143, view_149);  t_143 = view_149 = None\0A    t_144 = torch.ops.aten.t(mm_35);  mm_35 = None\0A    sum_43 = torch.ops.aten.sum(_reshape_alias_101, [0], True);  _reshape_alias_101 = None\0A    view_209 = torch.ops.aten.view(sum_43, [384]);  sum_43 = None\0A    detach_109 = torch.ops.aten.detach(view_209);  view_209 = None\0A    detach_110 = torch.ops.aten.detach(detach_109);  detach_109 = None\0A    _reshape_alias_102 = torch.ops.aten._reshape_alias(mm_34, [1, 128, 384], [49152, 384, 1]);  mm_34 = None\0A    add_46 = torch.ops.aten.add(mul_40, _reshape_alias_102);  mul_40 = _reshape_alias_102 = None\0A    t_145 = torch.ops.aten.t(t_144);  t_144 = None\0A    detach_111 = torch.ops.aten.detach(t_145);  t_145 = None\0A    detach_112 = torch.ops.aten.detach(detach_111);  detach_111 = None\0A    permute_59 = torch.ops.aten.permute(transpose_26, [0, 2, 1, 3]);  transpose_26 = None\0A    _reshape_alias_103 = torch.ops.aten._reshape_alias(permute_59, [1, 128, 384], [128, 1, 128]);  permute_59 = None\0A    _reshape_alias_104 = torch.ops.aten._reshape_alias(_reshape_alias_103, [128, 384], [1, 128]);  _reshape_alias_103 = None\0A    t_146 = torch.ops.aten.t(t_55);  t_55 = None\0A    mm_36 = torch.ops.aten.mm(_reshape_alias_104, t_146);  t_146 = None\0A    t_147 = torch.ops.aten.t(_reshape_alias_104)\0A    mm_37 = torch.ops.aten.mm(t_147, view_146);  t_147 = view_146 = None\0A    t_148 = torch.ops.aten.t(mm_37);  mm_37 = None\0A    sum_44 = torch.ops.aten.sum(_reshape_alias_104, [0], True);  _reshape_alias_104 = None\0A    view_210 = torch.ops.aten.view(sum_44, [384]);  sum_44 = None\0A    detach_113 = torch.ops.aten.detach(view_210);  view_210 = None\0A    detach_114 = torch.ops.aten.detach(detach_113);  detach_113 = None\0A    _reshape_alias_105 = torch.ops.aten._reshape_alias(mm_36, [1, 128, 384], [49152, 384, 1]);  mm_36 = None\0A    add_47 = torch.ops.aten.add(add_46, _reshape_alias_105);  add_46 = _reshape_alias_105 = None\0A    t_149 = torch.ops.aten.t(t_148);  t_148 = None\0A    detach_115 = torch.ops.aten.detach(t_149);  t_149 = None\0A    detach_116 = torch.ops.aten.detach(detach_115);  detach_115 = None\0A    _reshape_alias_106 = torch.ops.aten._reshape_alias(_unsafe_view_28, [128, 384], [384, 1]);  _unsafe_view_28 = None\0A    t_150 = torch.ops.aten.t(t_54);  t_54 = None\0A    mm_38 = torch.ops.aten.mm(_reshape_alias_106, t_150);  t_150 = None\0A    t_151 = torch.ops.aten.t(_reshape_alias_106)\0A    mm_39 = torch.ops.aten.mm(t_151, view_144);  t_151 = view_144 = None\0A    t_152 = torch.ops.aten.t(mm_39);  mm_39 = None\0A    sum_45 = torch.ops.aten.sum(_reshape_alias_106, [0], True);  _reshape_alias_106 = None\0A    view_211 = torch.ops.aten.view(sum_45, [384]);  sum_45 = None\0A    detach_117 = torch.ops.aten.detach(view_211);  view_211 = None\0A    detach_118 = torch.ops.aten.detach(detach_117);  detach_117 = None\0A    _reshape_alias_107 = torch.ops.aten._reshape_alias(mm_38, [1, 128, 384], [49152, 384, 1]);  mm_38 = None\0A    add_48 = torch.ops.aten.add(add_47, _reshape_alias_107);  add_47 = _reshape_alias_107 = None\0A    t_153 = torch.ops.aten.t(t_152);  t_152 = None\0A    detach_119 = torch.ops.aten.detach(t_153);  t_153 = None\0A    detach_120 = torch.ops.aten.detach(detach_119);  detach_119 = None\0A    to_54 = torch.ops.aten.to(add_48, 6)\0A    to_55 = torch.ops.aten.to(add_27, 6)\0A    to_56 = torch.ops.aten.to(getitem_55, 6)\0A    to_57 = torch.ops.aten.to(getitem_56, 6)\0A    to_58 = torch.ops.aten.to(params_179, 6)\0A    to_59 = torch.ops.aten.to(params_178, 6)\0A    sub_18 = torch.ops.aten.sub(add_27, getitem_55);  add_27 = getitem_55 = None\0A    mul_42 = torch.ops.aten.mul(sub_18, getitem_56);  sub_18 = None\0A    mul_43 = torch.ops.aten.mul(add_48, params_179)\0A    mul_44 = torch.ops.aten.mul(mul_43, 384)\0A    sum_46 = torch.ops.aten.sum(mul_43, [2], True)\0A    mul_45 = torch.ops.aten.mul(mul_43, mul_42);  mul_43 = None\0A    sum_47 = torch.ops.aten.sum(mul_45, [2], True);  mul_45 = None\0A    mul_46 = torch.ops.aten.mul(mul_42, sum_47);  sum_47 = None\0A    sub_19 = torch.ops.aten.sub(mul_44, sum_46);  mul_44 = sum_46 = None\0A    sub_20 = torch.ops.aten.sub(sub_19, mul_46);  sub_19 = mul_46 = None\0A    div_21 = torch.ops.aten.div(getitem_56, 384);  getitem_56 = None\0A    mul_47 = torch.ops.aten.mul(div_21, sub_20);  div_21 = sub_20 = None\0A    mul_48 = torch.ops.aten.mul(add_48, mul_42);  mul_42 = None\0A    sum_48 = torch.ops.aten.sum(mul_48, [0, 1]);  mul_48 = None\0A    sum_49 = torch.ops.aten.sum(add_48, [0, 1]);  add_48 = None\0A    to_60 = torch.ops.aten.to(mul_47, 6)\0A    to_61 = torch.ops.aten.to(sum_48, 6)\0A    to_62 = torch.ops.aten.to(sum_49, 6)\0A    detach_121 = torch.ops.aten.detach(sum_48);  sum_48 = None\0A    detach_122 = torch.ops.aten.detach(detach_121);  detach_121 = None\0A    detach_123 = torch.ops.aten.detach(sum_49);  sum_49 = None\0A    detach_124 = torch.ops.aten.detach(detach_123);  detach_123 = None\0A    _reshape_alias_108 = torch.ops.aten._reshape_alias(mul_47, [128, 384], [384, 1])\0A    t_154 = torch.ops.aten.t(t_53);  t_53 = None\0A    mm_40 = torch.ops.aten.mm(_reshape_alias_108, t_154);  t_154 = None\0A    t_155 = torch.ops.aten.t(_reshape_alias_108)\0A    mm_41 = torch.ops.aten.mm(t_155, view_142);  t_155 = view_142 = None\0A    t_156 = torch.ops.aten.t(mm_41);  mm_41 = None\0A    sum_50 = torch.ops.aten.sum(_reshape_alias_108, [0], True);  _reshape_alias_108 = None\0A    view_212 = torch.ops.aten.view(sum_50, [384]);  sum_50 = None\0A    detach_125 = torch.ops.aten.detach(view_212);  view_212 = None\0A    detach_126 = torch.ops.aten.detach(detach_125);  detach_125 = None\0A    _reshape_alias_109 = torch.ops.aten._reshape_alias(mm_40, [1, 128, 1536], [196608, 1536, 1]);  mm_40 = None\0A    t_157 = torch.ops.aten.t(t_156);  t_156 = None\0A    detach_127 = torch.ops.aten.detach(t_157);  t_157 = None\0A    detach_128 = torch.ops.aten.detach(detach_127);  detach_127 = None\0A    gelu_backward_3 = torch.ops.aten.gelu_backward(_reshape_alias_109, view_141);  _reshape_alias_109 = view_141 = None\0A    _reshape_alias_110 = torch.ops.aten._reshape_alias(gelu_backward_3, [128, 1536], [1536, 1]);  gelu_backward_3 = None\0A    t_158 = torch.ops.aten.t(t_52);  t_52 = None\0A    mm_42 = torch.ops.aten.mm(_reshape_alias_110, t_158);  t_158 = None\0A    t_159 = torch.ops.aten.t(_reshape_alias_110)\0A    mm_43 = torch.ops.aten.mm(t_159, view_140);  t_159 = view_140 = None\0A    t_160 = torch.ops.aten.t(mm_43);  mm_43 = None\0A    sum_51 = torch.ops.aten.sum(_reshape_alias_110, [0], True);  _reshape_alias_110 = None\0A    view_213 = torch.ops.aten.view(sum_51, [1536]);  sum_51 = None\0A    detach_129 = torch.ops.aten.detach(view_213);  view_213 = None\0A    detach_130 = torch.ops.aten.detach(detach_129);  detach_129 = None\0A    _reshape_alias_111 = torch.ops.aten._reshape_alias(mm_42, [1, 128, 384], [49152, 384, 1]);  mm_42 = None\0A    add_49 = torch.ops.aten.add(mul_47, _reshape_alias_111);  mul_47 = _reshape_alias_111 = None\0A    t_161 = torch.ops.aten.t(t_160);  t_160 = None\0A    detach_131 = torch.ops.aten.detach(t_161);  t_161 = None\0A    detach_132 = torch.ops.aten.detach(detach_131);  detach_131 = None\0A    to_63 = torch.ops.aten.to(add_49, 6)\0A    to_64 = torch.ops.aten.to(add_26, 6)\0A    to_65 = torch.ops.aten.to(getitem_52, 6)\0A    to_66 = torch.ops.aten.to(getitem_53, 6)\0A    to_67 = torch.ops.aten.to(params_167, 6)\0A    to_68 = torch.ops.aten.to(params_166, 6)\0A    sub_21 = torch.ops.aten.sub(add_26, getitem_52);  add_26 = getitem_52 = None\0A    mul_49 = torch.ops.aten.mul(sub_21, getitem_53);  sub_21 = None\0A    mul_50 = torch.ops.aten.mul(add_49, params_167)\0A    mul_51 = torch.ops.aten.mul(mul_50, 384)\0A    sum_52 = torch.ops.aten.sum(mul_50, [2], True)\0A    mul_52 = torch.ops.aten.mul(mul_50, mul_49);  mul_50 = None\0A    sum_53 = torch.ops.aten.sum(mul_52, [2], True);  mul_52 = None\0A    mul_53 = torch.ops.aten.mul(mul_49, sum_53);  sum_53 = None\0A    sub_22 = torch.ops.aten.sub(mul_51, sum_52);  mul_51 = sum_52 = None\0A    sub_23 = torch.ops.aten.sub(sub_22, mul_53);  sub_22 = mul_53 = None\0A    div_22 = torch.ops.aten.div(getitem_53, 384);  getitem_53 = None\0A    mul_54 = torch.ops.aten.mul(div_22, sub_23);  div_22 = sub_23 = None\0A    mul_55 = torch.ops.aten.mul(add_49, mul_49);  mul_49 = None\0A    sum_54 = torch.ops.aten.sum(mul_55, [0, 1]);  mul_55 = None\0A    sum_55 = torch.ops.aten.sum(add_49, [0, 1]);  add_49 = None\0A    to_69 = torch.ops.aten.to(mul_54, 6)\0A    to_70 = torch.ops.aten.to(sum_54, 6)\0A    to_71 = torch.ops.aten.to(sum_55, 6)\0A    detach_133 = torch.ops.aten.detach(sum_54);  sum_54 = None\0A    detach_134 = torch.ops.aten.detach(detach_133);  detach_133 = None\0A    detach_135 = torch.ops.aten.detach(sum_55);  sum_55 = None\0A    detach_136 = torch.ops.aten.detach(detach_135);  detach_135 = None\0A    _reshape_alias_112 = torch.ops.aten._reshape_alias(mul_54, [128, 384], [384, 1])\0A    t_162 = torch.ops.aten.t(t_51);  t_51 = None\0A    mm_44 = torch.ops.aten.mm(_reshape_alias_112, t_162);  t_162 = None\0A    t_163 = torch.ops.aten.t(_reshape_alias_112)\0A    mm_45 = torch.ops.aten.mm(t_163, view_138);  t_163 = view_138 = None\0A    t_164 = torch.ops.aten.t(mm_45);  mm_45 = None\0A    sum_56 = torch.ops.aten.sum(_reshape_alias_112, [0], True);  _reshape_alias_112 = None\0A    view_214 = torch.ops.aten.view(sum_56, [384]);  sum_56 = None\0A    detach_137 = torch.ops.aten.detach(view_214);  view_214 = None\0A    detach_138 = torch.ops.aten.detach(detach_137);  detach_137 = None\0A    _reshape_alias_113 = torch.ops.aten._reshape_alias(mm_44, [1, 128, 384], [49152, 384, 1]);  mm_44 = None\0A    t_165 = torch.ops.aten.t(t_164);  t_164 = None\0A    detach_139 = torch.ops.aten.detach(t_165);  t_165 = None\0A    detach_140 = torch.ops.aten.detach(detach_139);  detach_139 = None\0A    _reshape_alias_114 = torch.ops.aten._reshape_alias(_reshape_alias_113, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_113 = None\0A    permute_60 = torch.ops.aten.permute(_reshape_alias_114, [0, 2, 1, 3]);  _reshape_alias_114 = None\0A    _reshape_alias_115 = torch.ops.aten._reshape_alias(permute_60, [12, 128, 32], [32, 384, 1]);  permute_60 = None\0A    transpose_27 = torch.ops.aten.transpose(_reshape_alias_34, 1, 2);  _reshape_alias_34 = None\0A    bmm_36 = torch.ops.aten.bmm(transpose_27, _reshape_alias_115);  transpose_27 = None\0A    transpose_28 = torch.ops.aten.transpose(_reshape_alias_35, 1, 2);  _reshape_alias_35 = None\0A    bmm_37 = torch.ops.aten.bmm(_reshape_alias_115, transpose_28);  _reshape_alias_115 = transpose_28 = None\0A    _reshape_alias_116 = torch.ops.aten._reshape_alias(bmm_36, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_36 = None\0A    _reshape_alias_117 = torch.ops.aten._reshape_alias(bmm_37, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_37 = None\0A    detach_141 = torch.ops.aten.detach(detach_8);  detach_8 = None\0A    _softmax_backward_data_3 = torch.ops.aten._softmax_backward_data(_reshape_alias_117, detach_141, -1, 6);  _reshape_alias_117 = detach_141 = None\0A    div_23 = torch.ops.aten.div(_softmax_backward_data_3, 5.656854249492381);  _softmax_backward_data_3 = None\0A    _reshape_alias_118 = torch.ops.aten._reshape_alias(div_23, [12, 128, 128], [16384, 128, 1]);  div_23 = None\0A    transpose_29 = torch.ops.aten.transpose(_reshape_alias_32, 1, 2);  _reshape_alias_32 = None\0A    bmm_38 = torch.ops.aten.bmm(transpose_29, _reshape_alias_118);  transpose_29 = None\0A    transpose_30 = torch.ops.aten.transpose(_reshape_alias_33, 1, 2);  _reshape_alias_33 = None\0A    bmm_39 = torch.ops.aten.bmm(_reshape_alias_118, transpose_30);  _reshape_alias_118 = transpose_30 = None\0A    _reshape_alias_119 = torch.ops.aten._reshape_alias(bmm_38, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_38 = None\0A    _reshape_alias_120 = torch.ops.aten._reshape_alias(bmm_39, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_39 = None\0A    transpose_31 = torch.ops.aten.transpose(_reshape_alias_119, -1, -2);  _reshape_alias_119 = None\0A    permute_61 = torch.ops.aten.permute(_reshape_alias_120, [0, 2, 1, 3]);  _reshape_alias_120 = None\0A    clone_18 = torch.ops.aten.clone(permute_61, memory_format = 0);  permute_61 = None\0A    _unsafe_view_30 = torch.ops.aten._unsafe_view(clone_18, [1, 128, 384]);  clone_18 = None\0A    permute_62 = torch.ops.aten.permute(_reshape_alias_116, [0, 2, 1, 3]);  _reshape_alias_116 = None\0A    clone_19 = torch.ops.aten.clone(permute_62, memory_format = 0);  permute_62 = None\0A    _unsafe_view_31 = torch.ops.aten._unsafe_view(clone_19, [1, 128, 384]);  clone_19 = None\0A    _reshape_alias_121 = torch.ops.aten._reshape_alias(_unsafe_view_31, [128, 384], [384, 1]);  _unsafe_view_31 = None\0A    t_166 = torch.ops.aten.t(t_50);  t_50 = None\0A    mm_46 = torch.ops.aten.mm(_reshape_alias_121, t_166);  t_166 = None\0A    t_167 = torch.ops.aten.t(_reshape_alias_121)\0A    mm_47 = torch.ops.aten.mm(t_167, view_133);  t_167 = view_133 = None\0A    t_168 = torch.ops.aten.t(mm_47);  mm_47 = None\0A    sum_57 = torch.ops.aten.sum(_reshape_alias_121, [0], True);  _reshape_alias_121 = None\0A    view_215 = torch.ops.aten.view(sum_57, [384]);  sum_57 = None\0A    detach_142 = torch.ops.aten.detach(view_215);  view_215 = None\0A    detach_143 = torch.ops.aten.detach(detach_142);  detach_142 = None\0A    _reshape_alias_122 = torch.ops.aten._reshape_alias(mm_46, [1, 128, 384], [49152, 384, 1]);  mm_46 = None\0A    add_50 = torch.ops.aten.add(mul_54, _reshape_alias_122);  mul_54 = _reshape_alias_122 = None\0A    t_169 = torch.ops.aten.t(t_168);  t_168 = None\0A    detach_144 = torch.ops.aten.detach(t_169);  t_169 = None\0A    detach_145 = torch.ops.aten.detach(detach_144);  detach_144 = None\0A    permute_63 = torch.ops.aten.permute(transpose_31, [0, 2, 1, 3]);  transpose_31 = None\0A    _reshape_alias_123 = torch.ops.aten._reshape_alias(permute_63, [1, 128, 384], [128, 1, 128]);  permute_63 = None\0A    _reshape_alias_124 = torch.ops.aten._reshape_alias(_reshape_alias_123, [128, 384], [1, 128]);  _reshape_alias_123 = None\0A    t_170 = torch.ops.aten.t(t_49);  t_49 = None\0A    mm_48 = torch.ops.aten.mm(_reshape_alias_124, t_170);  t_170 = None\0A    t_171 = torch.ops.aten.t(_reshape_alias_124)\0A    mm_49 = torch.ops.aten.mm(t_171, view_130);  t_171 = view_130 = None\0A    t_172 = torch.ops.aten.t(mm_49);  mm_49 = None\0A    sum_58 = torch.ops.aten.sum(_reshape_alias_124, [0], True);  _reshape_alias_124 = None\0A    view_216 = torch.ops.aten.view(sum_58, [384]);  sum_58 = None\0A    detach_146 = torch.ops.aten.detach(view_216);  view_216 = None\0A    detach_147 = torch.ops.aten.detach(detach_146);  detach_146 = None\0A    _reshape_alias_125 = torch.ops.aten._reshape_alias(mm_48, [1, 128, 384], [49152, 384, 1]);  mm_48 = None\0A    add_51 = torch.ops.aten.add(add_50, _reshape_alias_125);  add_50 = _reshape_alias_125 = None\0A    t_173 = torch.ops.aten.t(t_172);  t_172 = None\0A    detach_148 = torch.ops.aten.detach(t_173);  t_173 = None\0A    detach_149 = torch.ops.aten.detach(detach_148);  detach_148 = None\0A    _reshape_alias_126 = torch.ops.aten._reshape_alias(_unsafe_view_30, [128, 384], [384, 1]);  _unsafe_view_30 = None\0A    t_174 = torch.ops.aten.t(t_48);  t_48 = None\0A    mm_50 = torch.ops.aten.mm(_reshape_alias_126, t_174);  t_174 = None\0A    t_175 = torch.ops.aten.t(_reshape_alias_126)\0A    mm_51 = torch.ops.aten.mm(t_175, view_128);  t_175 = view_128 = None\0A    t_176 = torch.ops.aten.t(mm_51);  mm_51 = None\0A    sum_59 = torch.ops.aten.sum(_reshape_alias_126, [0], True);  _reshape_alias_126 = None\0A    view_217 = torch.ops.aten.view(sum_59, [384]);  sum_59 = None\0A    detach_150 = torch.ops.aten.detach(view_217);  view_217 = None\0A    detach_151 = torch.ops.aten.detach(detach_150);  detach_150 = None\0A    _reshape_alias_127 = torch.ops.aten._reshape_alias(mm_50, [1, 128, 384], [49152, 384, 1]);  mm_50 = None\0A    add_52 = torch.ops.aten.add(add_51, _reshape_alias_127);  add_51 = _reshape_alias_127 = None\0A    t_177 = torch.ops.aten.t(t_176);  t_176 = None\0A    detach_152 = torch.ops.aten.detach(t_177);  t_177 = None\0A    detach_153 = torch.ops.aten.detach(detach_152);  detach_152 = None\0A    to_72 = torch.ops.aten.to(add_52, 6)\0A    to_73 = torch.ops.aten.to(add_24, 6)\0A    to_74 = torch.ops.aten.to(getitem_49, 6)\0A    to_75 = torch.ops.aten.to(getitem_50, 6)\0A    to_76 = torch.ops.aten.to(params_163, 6)\0A    to_77 = torch.ops.aten.to(params_162, 6)\0A    sub_24 = torch.ops.aten.sub(add_24, getitem_49);  add_24 = getitem_49 = None\0A    mul_56 = torch.ops.aten.mul(sub_24, getitem_50);  sub_24 = None\0A    mul_57 = torch.ops.aten.mul(add_52, params_163)\0A    mul_58 = torch.ops.aten.mul(mul_57, 384)\0A    sum_60 = torch.ops.aten.sum(mul_57, [2], True)\0A    mul_59 = torch.ops.aten.mul(mul_57, mul_56);  mul_57 = None\0A    sum_61 = torch.ops.aten.sum(mul_59, [2], True);  mul_59 = None\0A    mul_60 = torch.ops.aten.mul(mul_56, sum_61);  sum_61 = None\0A    sub_25 = torch.ops.aten.sub(mul_58, sum_60);  mul_58 = sum_60 = None\0A    sub_26 = torch.ops.aten.sub(sub_25, mul_60);  sub_25 = mul_60 = None\0A    div_24 = torch.ops.aten.div(getitem_50, 384);  getitem_50 = None\0A    mul_61 = torch.ops.aten.mul(div_24, sub_26);  div_24 = sub_26 = None\0A    mul_62 = torch.ops.aten.mul(add_52, mul_56);  mul_56 = None\0A    sum_62 = torch.ops.aten.sum(mul_62, [0, 1]);  mul_62 = None\0A    sum_63 = torch.ops.aten.sum(add_52, [0, 1]);  add_52 = None\0A    to_78 = torch.ops.aten.to(mul_61, 6)\0A    to_79 = torch.ops.aten.to(sum_62, 6)\0A    to_80 = torch.ops.aten.to(sum_63, 6)\0A    detach_154 = torch.ops.aten.detach(sum_62);  sum_62 = None\0A    detach_155 = torch.ops.aten.detach(detach_154);  detach_154 = None\0A    detach_156 = torch.ops.aten.detach(sum_63);  sum_63 = None\0A    detach_157 = torch.ops.aten.detach(detach_156);  detach_156 = None\0A    _reshape_alias_128 = torch.ops.aten._reshape_alias(mul_61, [128, 384], [384, 1])\0A    t_178 = torch.ops.aten.t(t_47);  t_47 = None\0A    mm_52 = torch.ops.aten.mm(_reshape_alias_128, t_178);  t_178 = None\0A    t_179 = torch.ops.aten.t(_reshape_alias_128)\0A    mm_53 = torch.ops.aten.mm(t_179, view_126);  t_179 = view_126 = None\0A    t_180 = torch.ops.aten.t(mm_53);  mm_53 = None\0A    sum_64 = torch.ops.aten.sum(_reshape_alias_128, [0], True);  _reshape_alias_128 = None\0A    view_218 = torch.ops.aten.view(sum_64, [384]);  sum_64 = None\0A    detach_158 = torch.ops.aten.detach(view_218);  view_218 = None\0A    detach_159 = torch.ops.aten.detach(detach_158);  detach_158 = None\0A    _reshape_alias_129 = torch.ops.aten._reshape_alias(mm_52, [1, 128, 1536], [196608, 1536, 1]);  mm_52 = None\0A    t_181 = torch.ops.aten.t(t_180);  t_180 = None\0A    detach_160 = torch.ops.aten.detach(t_181);  t_181 = None\0A    detach_161 = torch.ops.aten.detach(detach_160);  detach_160 = None\0A    gelu_backward_4 = torch.ops.aten.gelu_backward(_reshape_alias_129, view_125);  _reshape_alias_129 = view_125 = None\0A    _reshape_alias_130 = torch.ops.aten._reshape_alias(gelu_backward_4, [128, 1536], [1536, 1]);  gelu_backward_4 = None\0A    t_182 = torch.ops.aten.t(t_46);  t_46 = None\0A    mm_54 = torch.ops.aten.mm(_reshape_alias_130, t_182);  t_182 = None\0A    t_183 = torch.ops.aten.t(_reshape_alias_130)\0A    mm_55 = torch.ops.aten.mm(t_183, view_124);  t_183 = view_124 = None\0A    t_184 = torch.ops.aten.t(mm_55);  mm_55 = None\0A    sum_65 = torch.ops.aten.sum(_reshape_alias_130, [0], True);  _reshape_alias_130 = None\0A    view_219 = torch.ops.aten.view(sum_65, [1536]);  sum_65 = None\0A    detach_162 = torch.ops.aten.detach(view_219);  view_219 = None\0A    detach_163 = torch.ops.aten.detach(detach_162);  detach_162 = None\0A    _reshape_alias_131 = torch.ops.aten._reshape_alias(mm_54, [1, 128, 384], [49152, 384, 1]);  mm_54 = None\0A    add_53 = torch.ops.aten.add(mul_61, _reshape_alias_131);  mul_61 = _reshape_alias_131 = None\0A    t_185 = torch.ops.aten.t(t_184);  t_184 = None\0A    detach_164 = torch.ops.aten.detach(t_185);  t_185 = None\0A    detach_165 = torch.ops.aten.detach(detach_164);  detach_164 = None\0A    to_81 = torch.ops.aten.to(add_53, 6)\0A    to_82 = torch.ops.aten.to(add_23, 6)\0A    to_83 = torch.ops.aten.to(getitem_46, 6)\0A    to_84 = torch.ops.aten.to(getitem_47, 6)\0A    to_85 = torch.ops.aten.to(params_151, 6)\0A    to_86 = torch.ops.aten.to(params_150, 6)\0A    sub_27 = torch.ops.aten.sub(add_23, getitem_46);  add_23 = getitem_46 = None\0A    mul_63 = torch.ops.aten.mul(sub_27, getitem_47);  sub_27 = None\0A    mul_64 = torch.ops.aten.mul(add_53, params_151)\0A    mul_65 = torch.ops.aten.mul(mul_64, 384)\0A    sum_66 = torch.ops.aten.sum(mul_64, [2], True)\0A    mul_66 = torch.ops.aten.mul(mul_64, mul_63);  mul_64 = None\0A    sum_67 = torch.ops.aten.sum(mul_66, [2], True);  mul_66 = None\0A    mul_67 = torch.ops.aten.mul(mul_63, sum_67);  sum_67 = None\0A    sub_28 = torch.ops.aten.sub(mul_65, sum_66);  mul_65 = sum_66 = None\0A    sub_29 = torch.ops.aten.sub(sub_28, mul_67);  sub_28 = mul_67 = None\0A    div_25 = torch.ops.aten.div(getitem_47, 384);  getitem_47 = None\0A    mul_68 = torch.ops.aten.mul(div_25, sub_29);  div_25 = sub_29 = None\0A    mul_69 = torch.ops.aten.mul(add_53, mul_63);  mul_63 = None\0A    sum_68 = torch.ops.aten.sum(mul_69, [0, 1]);  mul_69 = None\0A    sum_69 = torch.ops.aten.sum(add_53, [0, 1]);  add_53 = None\0A    to_87 = torch.ops.aten.to(mul_68, 6)\0A    to_88 = torch.ops.aten.to(sum_68, 6)\0A    to_89 = torch.ops.aten.to(sum_69, 6)\0A    detach_166 = torch.ops.aten.detach(sum_68);  sum_68 = None\0A    detach_167 = torch.ops.aten.detach(detach_166);  detach_166 = None\0A    detach_168 = torch.ops.aten.detach(sum_69);  sum_69 = None\0A    detach_169 = torch.ops.aten.detach(detach_168);  detach_168 = None\0A    _reshape_alias_132 = torch.ops.aten._reshape_alias(mul_68, [128, 384], [384, 1])\0A    t_186 = torch.ops.aten.t(t_45);  t_45 = None\0A    mm_56 = torch.ops.aten.mm(_reshape_alias_132, t_186);  t_186 = None\0A    t_187 = torch.ops.aten.t(_reshape_alias_132)\0A    mm_57 = torch.ops.aten.mm(t_187, view_122);  t_187 = view_122 = None\0A    t_188 = torch.ops.aten.t(mm_57);  mm_57 = None\0A    sum_70 = torch.ops.aten.sum(_reshape_alias_132, [0], True);  _reshape_alias_132 = None\0A    view_220 = torch.ops.aten.view(sum_70, [384]);  sum_70 = None\0A    detach_170 = torch.ops.aten.detach(view_220);  view_220 = None\0A    detach_171 = torch.ops.aten.detach(detach_170);  detach_170 = None\0A    _reshape_alias_133 = torch.ops.aten._reshape_alias(mm_56, [1, 128, 384], [49152, 384, 1]);  mm_56 = None\0A    t_189 = torch.ops.aten.t(t_188);  t_188 = None\0A    detach_172 = torch.ops.aten.detach(t_189);  t_189 = None\0A    detach_173 = torch.ops.aten.detach(detach_172);  detach_172 = None\0A    _reshape_alias_134 = torch.ops.aten._reshape_alias(_reshape_alias_133, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_133 = None\0A    permute_64 = torch.ops.aten.permute(_reshape_alias_134, [0, 2, 1, 3]);  _reshape_alias_134 = None\0A    _reshape_alias_135 = torch.ops.aten._reshape_alias(permute_64, [12, 128, 32], [32, 384, 1]);  permute_64 = None\0A    transpose_32 = torch.ops.aten.transpose(_reshape_alias_30, 1, 2);  _reshape_alias_30 = None\0A    bmm_40 = torch.ops.aten.bmm(transpose_32, _reshape_alias_135);  transpose_32 = None\0A    transpose_33 = torch.ops.aten.transpose(_reshape_alias_31, 1, 2);  _reshape_alias_31 = None\0A    bmm_41 = torch.ops.aten.bmm(_reshape_alias_135, transpose_33);  _reshape_alias_135 = transpose_33 = None\0A    _reshape_alias_136 = torch.ops.aten._reshape_alias(bmm_40, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_40 = None\0A    _reshape_alias_137 = torch.ops.aten._reshape_alias(bmm_41, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_41 = None\0A    detach_174 = torch.ops.aten.detach(detach_7);  detach_7 = None\0A    _softmax_backward_data_4 = torch.ops.aten._softmax_backward_data(_reshape_alias_137, detach_174, -1, 6);  _reshape_alias_137 = detach_174 = None\0A    div_26 = torch.ops.aten.div(_softmax_backward_data_4, 5.656854249492381);  _softmax_backward_data_4 = None\0A    _reshape_alias_138 = torch.ops.aten._reshape_alias(div_26, [12, 128, 128], [16384, 128, 1]);  div_26 = None\0A    transpose_34 = torch.ops.aten.transpose(_reshape_alias_28, 1, 2);  _reshape_alias_28 = None\0A    bmm_42 = torch.ops.aten.bmm(transpose_34, _reshape_alias_138);  transpose_34 = None\0A    transpose_35 = torch.ops.aten.transpose(_reshape_alias_29, 1, 2);  _reshape_alias_29 = None\0A    bmm_43 = torch.ops.aten.bmm(_reshape_alias_138, transpose_35);  _reshape_alias_138 = transpose_35 = None\0A    _reshape_alias_139 = torch.ops.aten._reshape_alias(bmm_42, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_42 = None\0A    _reshape_alias_140 = torch.ops.aten._reshape_alias(bmm_43, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_43 = None\0A    transpose_36 = torch.ops.aten.transpose(_reshape_alias_139, -1, -2);  _reshape_alias_139 = None\0A    permute_65 = torch.ops.aten.permute(_reshape_alias_140, [0, 2, 1, 3]);  _reshape_alias_140 = None\0A    clone_20 = torch.ops.aten.clone(permute_65, memory_format = 0);  permute_65 = None\0A    _unsafe_view_32 = torch.ops.aten._unsafe_view(clone_20, [1, 128, 384]);  clone_20 = None\0A    permute_66 = torch.ops.aten.permute(_reshape_alias_136, [0, 2, 1, 3]);  _reshape_alias_136 = None\0A    clone_21 = torch.ops.aten.clone(permute_66, memory_format = 0);  permute_66 = None\0A    _unsafe_view_33 = torch.ops.aten._unsafe_view(clone_21, [1, 128, 384]);  clone_21 = None\0A    _reshape_alias_141 = torch.ops.aten._reshape_alias(_unsafe_view_33, [128, 384], [384, 1]);  _unsafe_view_33 = None\0A    t_190 = torch.ops.aten.t(t_44);  t_44 = None\0A    mm_58 = torch.ops.aten.mm(_reshape_alias_141, t_190);  t_190 = None\0A    t_191 = torch.ops.aten.t(_reshape_alias_141)\0A    mm_59 = torch.ops.aten.mm(t_191, view_117);  t_191 = view_117 = None\0A    t_192 = torch.ops.aten.t(mm_59);  mm_59 = None\0A    sum_71 = torch.ops.aten.sum(_reshape_alias_141, [0], True);  _reshape_alias_141 = None\0A    view_221 = torch.ops.aten.view(sum_71, [384]);  sum_71 = None\0A    detach_175 = torch.ops.aten.detach(view_221);  view_221 = None\0A    detach_176 = torch.ops.aten.detach(detach_175);  detach_175 = None\0A    _reshape_alias_142 = torch.ops.aten._reshape_alias(mm_58, [1, 128, 384], [49152, 384, 1]);  mm_58 = None\0A    add_54 = torch.ops.aten.add(mul_68, _reshape_alias_142);  mul_68 = _reshape_alias_142 = None\0A    t_193 = torch.ops.aten.t(t_192);  t_192 = None\0A    detach_177 = torch.ops.aten.detach(t_193);  t_193 = None\0A    detach_178 = torch.ops.aten.detach(detach_177);  detach_177 = None\0A    permute_67 = torch.ops.aten.permute(transpose_36, [0, 2, 1, 3]);  transpose_36 = None\0A    _reshape_alias_143 = torch.ops.aten._reshape_alias(permute_67, [1, 128, 384], [128, 1, 128]);  permute_67 = None\0A    _reshape_alias_144 = torch.ops.aten._reshape_alias(_reshape_alias_143, [128, 384], [1, 128]);  _reshape_alias_143 = None\0A    t_194 = torch.ops.aten.t(t_43);  t_43 = None\0A    mm_60 = torch.ops.aten.mm(_reshape_alias_144, t_194);  t_194 = None\0A    t_195 = torch.ops.aten.t(_reshape_alias_144)\0A    mm_61 = torch.ops.aten.mm(t_195, view_114);  t_195 = view_114 = None\0A    t_196 = torch.ops.aten.t(mm_61);  mm_61 = None\0A    sum_72 = torch.ops.aten.sum(_reshape_alias_144, [0], True);  _reshape_alias_144 = None\0A    view_222 = torch.ops.aten.view(sum_72, [384]);  sum_72 = None\0A    detach_179 = torch.ops.aten.detach(view_222);  view_222 = None\0A    detach_180 = torch.ops.aten.detach(detach_179);  detach_179 = None\0A    _reshape_alias_145 = torch.ops.aten._reshape_alias(mm_60, [1, 128, 384], [49152, 384, 1]);  mm_60 = None\0A    add_55 = torch.ops.aten.add(add_54, _reshape_alias_145);  add_54 = _reshape_alias_145 = None\0A    t_197 = torch.ops.aten.t(t_196);  t_196 = None\0A    detach_181 = torch.ops.aten.detach(t_197);  t_197 = None\0A    detach_182 = torch.ops.aten.detach(detach_181);  detach_181 = None\0A    _reshape_alias_146 = torch.ops.aten._reshape_alias(_unsafe_view_32, [128, 384], [384, 1]);  _unsafe_view_32 = None\0A    t_198 = torch.ops.aten.t(t_42);  t_42 = None\0A    mm_62 = torch.ops.aten.mm(_reshape_alias_146, t_198);  t_198 = None\0A    t_199 = torch.ops.aten.t(_reshape_alias_146)\0A    mm_63 = torch.ops.aten.mm(t_199, view_112);  t_199 = view_112 = None\0A    t_200 = torch.ops.aten.t(mm_63);  mm_63 = None\0A    sum_73 = torch.ops.aten.sum(_reshape_alias_146, [0], True);  _reshape_alias_146 = None\0A    view_223 = torch.ops.aten.view(sum_73, [384]);  sum_73 = None\0A    detach_183 = torch.ops.aten.detach(view_223);  view_223 = None\0A    detach_184 = torch.ops.aten.detach(detach_183);  detach_183 = None\0A    _reshape_alias_147 = torch.ops.aten._reshape_alias(mm_62, [1, 128, 384], [49152, 384, 1]);  mm_62 = None\0A    add_56 = torch.ops.aten.add(add_55, _reshape_alias_147);  add_55 = _reshape_alias_147 = None\0A    t_201 = torch.ops.aten.t(t_200);  t_200 = None\0A    detach_185 = torch.ops.aten.detach(t_201);  t_201 = None\0A    detach_186 = torch.ops.aten.detach(detach_185);  detach_185 = None\0A    to_90 = torch.ops.aten.to(add_56, 6)\0A    to_91 = torch.ops.aten.to(add_21, 6)\0A    to_92 = torch.ops.aten.to(getitem_43, 6)\0A    to_93 = torch.ops.aten.to(getitem_44, 6)\0A    to_94 = torch.ops.aten.to(params_147, 6)\0A    to_95 = torch.ops.aten.to(params_146, 6)\0A    sub_30 = torch.ops.aten.sub(add_21, getitem_43);  add_21 = getitem_43 = None\0A    mul_70 = torch.ops.aten.mul(sub_30, getitem_44);  sub_30 = None\0A    mul_71 = torch.ops.aten.mul(add_56, params_147)\0A    mul_72 = torch.ops.aten.mul(mul_71, 384)\0A    sum_74 = torch.ops.aten.sum(mul_71, [2], True)\0A    mul_73 = torch.ops.aten.mul(mul_71, mul_70);  mul_71 = None\0A    sum_75 = torch.ops.aten.sum(mul_73, [2], True);  mul_73 = None\0A    mul_74 = torch.ops.aten.mul(mul_70, sum_75);  sum_75 = None\0A    sub_31 = torch.ops.aten.sub(mul_72, sum_74);  mul_72 = sum_74 = None\0A    sub_32 = torch.ops.aten.sub(sub_31, mul_74);  sub_31 = mul_74 = None\0A    div_27 = torch.ops.aten.div(getitem_44, 384);  getitem_44 = None\0A    mul_75 = torch.ops.aten.mul(div_27, sub_32);  div_27 = sub_32 = None\0A    mul_76 = torch.ops.aten.mul(add_56, mul_70);  mul_70 = None\0A    sum_76 = torch.ops.aten.sum(mul_76, [0, 1]);  mul_76 = None\0A    sum_77 = torch.ops.aten.sum(add_56, [0, 1]);  add_56 = None\0A    to_96 = torch.ops.aten.to(mul_75, 6)\0A    to_97 = torch.ops.aten.to(sum_76, 6)\0A    to_98 = torch.ops.aten.to(sum_77, 6)\0A    detach_187 = torch.ops.aten.detach(sum_76);  sum_76 = None\0A    detach_188 = torch.ops.aten.detach(detach_187);  detach_187 = None\0A    detach_189 = torch.ops.aten.detach(sum_77);  sum_77 = None\0A    detach_190 = torch.ops.aten.detach(detach_189);  detach_189 = None\0A    _reshape_alias_148 = torch.ops.aten._reshape_alias(mul_75, [128, 384], [384, 1])\0A    t_202 = torch.ops.aten.t(t_41);  t_41 = None\0A    mm_64 = torch.ops.aten.mm(_reshape_alias_148, t_202);  t_202 = None\0A    t_203 = torch.ops.aten.t(_reshape_alias_148)\0A    mm_65 = torch.ops.aten.mm(t_203, view_110);  t_203 = view_110 = None\0A    t_204 = torch.ops.aten.t(mm_65);  mm_65 = None\0A    sum_78 = torch.ops.aten.sum(_reshape_alias_148, [0], True);  _reshape_alias_148 = None\0A    view_224 = torch.ops.aten.view(sum_78, [384]);  sum_78 = None\0A    detach_191 = torch.ops.aten.detach(view_224);  view_224 = None\0A    detach_192 = torch.ops.aten.detach(detach_191);  detach_191 = None\0A    _reshape_alias_149 = torch.ops.aten._reshape_alias(mm_64, [1, 128, 1536], [196608, 1536, 1]);  mm_64 = None\0A    t_205 = torch.ops.aten.t(t_204);  t_204 = None\0A    detach_193 = torch.ops.aten.detach(t_205);  t_205 = None\0A    detach_194 = torch.ops.aten.detach(detach_193);  detach_193 = None\0A    gelu_backward_5 = torch.ops.aten.gelu_backward(_reshape_alias_149, view_109);  _reshape_alias_149 = view_109 = None\0A    _reshape_alias_150 = torch.ops.aten._reshape_alias(gelu_backward_5, [128, 1536], [1536, 1]);  gelu_backward_5 = None\0A    t_206 = torch.ops.aten.t(t_40);  t_40 = None\0A    mm_66 = torch.ops.aten.mm(_reshape_alias_150, t_206);  t_206 = None\0A    t_207 = torch.ops.aten.t(_reshape_alias_150)\0A    mm_67 = torch.ops.aten.mm(t_207, view_108);  t_207 = view_108 = None\0A    t_208 = torch.ops.aten.t(mm_67);  mm_67 = None\0A    sum_79 = torch.ops.aten.sum(_reshape_alias_150, [0], True);  _reshape_alias_150 = None\0A    view_225 = torch.ops.aten.view(sum_79, [1536]);  sum_79 = None\0A    detach_195 = torch.ops.aten.detach(view_225);  view_225 = None\0A    detach_196 = torch.ops.aten.detach(detach_195);  detach_195 = None\0A    _reshape_alias_151 = torch.ops.aten._reshape_alias(mm_66, [1, 128, 384], [49152, 384, 1]);  mm_66 = None\0A    add_57 = torch.ops.aten.add(mul_75, _reshape_alias_151);  mul_75 = _reshape_alias_151 = None\0A    t_209 = torch.ops.aten.t(t_208);  t_208 = None\0A    detach_197 = torch.ops.aten.detach(t_209);  t_209 = None\0A    detach_198 = torch.ops.aten.detach(detach_197);  detach_197 = None\0A    to_99 = torch.ops.aten.to(add_57, 6)\0A    to_100 = torch.ops.aten.to(add_20, 6)\0A    to_101 = torch.ops.aten.to(getitem_40, 6)\0A    to_102 = torch.ops.aten.to(getitem_41, 6)\0A    to_103 = torch.ops.aten.to(params_135, 6)\0A    to_104 = torch.ops.aten.to(params_134, 6)\0A    sub_33 = torch.ops.aten.sub(add_20, getitem_40);  add_20 = getitem_40 = None\0A    mul_77 = torch.ops.aten.mul(sub_33, getitem_41);  sub_33 = None\0A    mul_78 = torch.ops.aten.mul(add_57, params_135)\0A    mul_79 = torch.ops.aten.mul(mul_78, 384)\0A    sum_80 = torch.ops.aten.sum(mul_78, [2], True)\0A    mul_80 = torch.ops.aten.mul(mul_78, mul_77);  mul_78 = None\0A    sum_81 = torch.ops.aten.sum(mul_80, [2], True);  mul_80 = None\0A    mul_81 = torch.ops.aten.mul(mul_77, sum_81);  sum_81 = None\0A    sub_34 = torch.ops.aten.sub(mul_79, sum_80);  mul_79 = sum_80 = None\0A    sub_35 = torch.ops.aten.sub(sub_34, mul_81);  sub_34 = mul_81 = None\0A    div_28 = torch.ops.aten.div(getitem_41, 384);  getitem_41 = None\0A    mul_82 = torch.ops.aten.mul(div_28, sub_35);  div_28 = sub_35 = None\0A    mul_83 = torch.ops.aten.mul(add_57, mul_77);  mul_77 = None\0A    sum_82 = torch.ops.aten.sum(mul_83, [0, 1]);  mul_83 = None\0A    sum_83 = torch.ops.aten.sum(add_57, [0, 1]);  add_57 = None\0A    to_105 = torch.ops.aten.to(mul_82, 6)\0A    to_106 = torch.ops.aten.to(sum_82, 6)\0A    to_107 = torch.ops.aten.to(sum_83, 6)\0A    detach_199 = torch.ops.aten.detach(sum_82);  sum_82 = None\0A    detach_200 = torch.ops.aten.detach(detach_199);  detach_199 = None\0A    detach_201 = torch.ops.aten.detach(sum_83);  sum_83 = None\0A    detach_202 = torch.ops.aten.detach(detach_201);  detach_201 = None\0A    _reshape_alias_152 = torch.ops.aten._reshape_alias(mul_82, [128, 384], [384, 1])\0A    t_210 = torch.ops.aten.t(t_39);  t_39 = None\0A    mm_68 = torch.ops.aten.mm(_reshape_alias_152, t_210);  t_210 = None\0A    t_211 = torch.ops.aten.t(_reshape_alias_152)\0A    mm_69 = torch.ops.aten.mm(t_211, view_106);  t_211 = view_106 = None\0A    t_212 = torch.ops.aten.t(mm_69);  mm_69 = None\0A    sum_84 = torch.ops.aten.sum(_reshape_alias_152, [0], True);  _reshape_alias_152 = None\0A    view_226 = torch.ops.aten.view(sum_84, [384]);  sum_84 = None\0A    detach_203 = torch.ops.aten.detach(view_226);  view_226 = None\0A    detach_204 = torch.ops.aten.detach(detach_203);  detach_203 = None\0A    _reshape_alias_153 = torch.ops.aten._reshape_alias(mm_68, [1, 128, 384], [49152, 384, 1]);  mm_68 = None\0A    t_213 = torch.ops.aten.t(t_212);  t_212 = None\0A    detach_205 = torch.ops.aten.detach(t_213);  t_213 = None\0A    detach_206 = torch.ops.aten.detach(detach_205);  detach_205 = None\0A    _reshape_alias_154 = torch.ops.aten._reshape_alias(_reshape_alias_153, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_153 = None\0A    permute_68 = torch.ops.aten.permute(_reshape_alias_154, [0, 2, 1, 3]);  _reshape_alias_154 = None\0A    _reshape_alias_155 = torch.ops.aten._reshape_alias(permute_68, [12, 128, 32], [32, 384, 1]);  permute_68 = None\0A    transpose_37 = torch.ops.aten.transpose(_reshape_alias_26, 1, 2);  _reshape_alias_26 = None\0A    bmm_44 = torch.ops.aten.bmm(transpose_37, _reshape_alias_155);  transpose_37 = None\0A    transpose_38 = torch.ops.aten.transpose(_reshape_alias_27, 1, 2);  _reshape_alias_27 = None\0A    bmm_45 = torch.ops.aten.bmm(_reshape_alias_155, transpose_38);  _reshape_alias_155 = transpose_38 = None\0A    _reshape_alias_156 = torch.ops.aten._reshape_alias(bmm_44, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_44 = None\0A    _reshape_alias_157 = torch.ops.aten._reshape_alias(bmm_45, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_45 = None\0A    detach_207 = torch.ops.aten.detach(detach_6);  detach_6 = None\0A    _softmax_backward_data_5 = torch.ops.aten._softmax_backward_data(_reshape_alias_157, detach_207, -1, 6);  _reshape_alias_157 = detach_207 = None\0A    div_29 = torch.ops.aten.div(_softmax_backward_data_5, 5.656854249492381);  _softmax_backward_data_5 = None\0A    _reshape_alias_158 = torch.ops.aten._reshape_alias(div_29, [12, 128, 128], [16384, 128, 1]);  div_29 = None\0A    transpose_39 = torch.ops.aten.transpose(_reshape_alias_24, 1, 2);  _reshape_alias_24 = None\0A    bmm_46 = torch.ops.aten.bmm(transpose_39, _reshape_alias_158);  transpose_39 = None\0A    transpose_40 = torch.ops.aten.transpose(_reshape_alias_25, 1, 2);  _reshape_alias_25 = None\0A    bmm_47 = torch.ops.aten.bmm(_reshape_alias_158, transpose_40);  _reshape_alias_158 = transpose_40 = None\0A    _reshape_alias_159 = torch.ops.aten._reshape_alias(bmm_46, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_46 = None\0A    _reshape_alias_160 = torch.ops.aten._reshape_alias(bmm_47, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_47 = None\0A    transpose_41 = torch.ops.aten.transpose(_reshape_alias_159, -1, -2);  _reshape_alias_159 = None\0A    permute_69 = torch.ops.aten.permute(_reshape_alias_160, [0, 2, 1, 3]);  _reshape_alias_160 = None\0A    clone_22 = torch.ops.aten.clone(permute_69, memory_format = 0);  permute_69 = None\0A    _unsafe_view_34 = torch.ops.aten._unsafe_view(clone_22, [1, 128, 384]);  clone_22 = None\0A    permute_70 = torch.ops.aten.permute(_reshape_alias_156, [0, 2, 1, 3]);  _reshape_alias_156 = None\0A    clone_23 = torch.ops.aten.clone(permute_70, memory_format = 0);  permute_70 = None\0A    _unsafe_view_35 = torch.ops.aten._unsafe_view(clone_23, [1, 128, 384]);  clone_23 = None\0A    _reshape_alias_161 = torch.ops.aten._reshape_alias(_unsafe_view_35, [128, 384], [384, 1]);  _unsafe_view_35 = None\0A    t_214 = torch.ops.aten.t(t_38);  t_38 = None\0A    mm_70 = torch.ops.aten.mm(_reshape_alias_161, t_214);  t_214 = None\0A    t_215 = torch.ops.aten.t(_reshape_alias_161)\0A    mm_71 = torch.ops.aten.mm(t_215, view_101);  t_215 = view_101 = None\0A    t_216 = torch.ops.aten.t(mm_71);  mm_71 = None\0A    sum_85 = torch.ops.aten.sum(_reshape_alias_161, [0], True);  _reshape_alias_161 = None\0A    view_227 = torch.ops.aten.view(sum_85, [384]);  sum_85 = None\0A    detach_208 = torch.ops.aten.detach(view_227);  view_227 = None\0A    detach_209 = torch.ops.aten.detach(detach_208);  detach_208 = None\0A    _reshape_alias_162 = torch.ops.aten._reshape_alias(mm_70, [1, 128, 384], [49152, 384, 1]);  mm_70 = None\0A    add_58 = torch.ops.aten.add(mul_82, _reshape_alias_162);  mul_82 = _reshape_alias_162 = None\0A    t_217 = torch.ops.aten.t(t_216);  t_216 = None\0A    detach_210 = torch.ops.aten.detach(t_217);  t_217 = None\0A    detach_211 = torch.ops.aten.detach(detach_210);  detach_210 = None\0A    permute_71 = torch.ops.aten.permute(transpose_41, [0, 2, 1, 3]);  transpose_41 = None\0A    _reshape_alias_163 = torch.ops.aten._reshape_alias(permute_71, [1, 128, 384], [128, 1, 128]);  permute_71 = None\0A    _reshape_alias_164 = torch.ops.aten._reshape_alias(_reshape_alias_163, [128, 384], [1, 128]);  _reshape_alias_163 = None\0A    t_218 = torch.ops.aten.t(t_37);  t_37 = None\0A    mm_72 = torch.ops.aten.mm(_reshape_alias_164, t_218);  t_218 = None\0A    t_219 = torch.ops.aten.t(_reshape_alias_164)\0A    mm_73 = torch.ops.aten.mm(t_219, view_98);  t_219 = view_98 = None\0A    t_220 = torch.ops.aten.t(mm_73);  mm_73 = None\0A    sum_86 = torch.ops.aten.sum(_reshape_alias_164, [0], True);  _reshape_alias_164 = None\0A    view_228 = torch.ops.aten.view(sum_86, [384]);  sum_86 = None\0A    detach_212 = torch.ops.aten.detach(view_228);  view_228 = None\0A    detach_213 = torch.ops.aten.detach(detach_212);  detach_212 = None\0A    _reshape_alias_165 = torch.ops.aten._reshape_alias(mm_72, [1, 128, 384], [49152, 384, 1]);  mm_72 = None\0A    add_59 = torch.ops.aten.add(add_58, _reshape_alias_165);  add_58 = _reshape_alias_165 = None\0A    t_221 = torch.ops.aten.t(t_220);  t_220 = None\0A    detach_214 = torch.ops.aten.detach(t_221);  t_221 = None\0A    detach_215 = torch.ops.aten.detach(detach_214);  detach_214 = None\0A    _reshape_alias_166 = torch.ops.aten._reshape_alias(_unsafe_view_34, [128, 384], [384, 1]);  _unsafe_view_34 = None\0A    t_222 = torch.ops.aten.t(t_36);  t_36 = None\0A    mm_74 = torch.ops.aten.mm(_reshape_alias_166, t_222);  t_222 = None\0A    t_223 = torch.ops.aten.t(_reshape_alias_166)\0A    mm_75 = torch.ops.aten.mm(t_223, view_96);  t_223 = view_96 = None\0A    t_224 = torch.ops.aten.t(mm_75);  mm_75 = None\0A    sum_87 = torch.ops.aten.sum(_reshape_alias_166, [0], True);  _reshape_alias_166 = None\0A    view_229 = torch.ops.aten.view(sum_87, [384]);  sum_87 = None\0A    detach_216 = torch.ops.aten.detach(view_229);  view_229 = None\0A    detach_217 = torch.ops.aten.detach(detach_216);  detach_216 = None\0A    _reshape_alias_167 = torch.ops.aten._reshape_alias(mm_74, [1, 128, 384], [49152, 384, 1]);  mm_74 = None\0A    add_60 = torch.ops.aten.add(add_59, _reshape_alias_167);  add_59 = _reshape_alias_167 = None\0A    t_225 = torch.ops.aten.t(t_224);  t_224 = None\0A    detach_218 = torch.ops.aten.detach(t_225);  t_225 = None\0A    detach_219 = torch.ops.aten.detach(detach_218);  detach_218 = None\0A    to_108 = torch.ops.aten.to(add_60, 6)\0A    to_109 = torch.ops.aten.to(add_18, 6)\0A    to_110 = torch.ops.aten.to(getitem_37, 6)\0A    to_111 = torch.ops.aten.to(getitem_38, 6)\0A    to_112 = torch.ops.aten.to(params_131, 6)\0A    to_113 = torch.ops.aten.to(params_130, 6)\0A    sub_36 = torch.ops.aten.sub(add_18, getitem_37);  add_18 = getitem_37 = None\0A    mul_84 = torch.ops.aten.mul(sub_36, getitem_38);  sub_36 = None\0A    mul_85 = torch.ops.aten.mul(add_60, params_131)\0A    mul_86 = torch.ops.aten.mul(mul_85, 384)\0A    sum_88 = torch.ops.aten.sum(mul_85, [2], True)\0A    mul_87 = torch.ops.aten.mul(mul_85, mul_84);  mul_85 = None\0A    sum_89 = torch.ops.aten.sum(mul_87, [2], True);  mul_87 = None\0A    mul_88 = torch.ops.aten.mul(mul_84, sum_89);  sum_89 = None\0A    sub_37 = torch.ops.aten.sub(mul_86, sum_88);  mul_86 = sum_88 = None\0A    sub_38 = torch.ops.aten.sub(sub_37, mul_88);  sub_37 = mul_88 = None\0A    div_30 = torch.ops.aten.div(getitem_38, 384);  getitem_38 = None\0A    mul_89 = torch.ops.aten.mul(div_30, sub_38);  div_30 = sub_38 = None\0A    mul_90 = torch.ops.aten.mul(add_60, mul_84);  mul_84 = None\0A    sum_90 = torch.ops.aten.sum(mul_90, [0, 1]);  mul_90 = None\0A    sum_91 = torch.ops.aten.sum(add_60, [0, 1]);  add_60 = None\0A    to_114 = torch.ops.aten.to(mul_89, 6)\0A    to_115 = torch.ops.aten.to(sum_90, 6)\0A    to_116 = torch.ops.aten.to(sum_91, 6)\0A    detach_220 = torch.ops.aten.detach(sum_90);  sum_90 = None\0A    detach_221 = torch.ops.aten.detach(detach_220);  detach_220 = None\0A    detach_222 = torch.ops.aten.detach(sum_91);  sum_91 = None\0A    detach_223 = torch.ops.aten.detach(detach_222);  detach_222 = None\0A    _reshape_alias_168 = torch.ops.aten._reshape_alias(mul_89, [128, 384], [384, 1])\0A    t_226 = torch.ops.aten.t(t_35);  t_35 = None\0A    mm_76 = torch.ops.aten.mm(_reshape_alias_168, t_226);  t_226 = None\0A    t_227 = torch.ops.aten.t(_reshape_alias_168)\0A    mm_77 = torch.ops.aten.mm(t_227, view_94);  t_227 = view_94 = None\0A    t_228 = torch.ops.aten.t(mm_77);  mm_77 = None\0A    sum_92 = torch.ops.aten.sum(_reshape_alias_168, [0], True);  _reshape_alias_168 = None\0A    view_230 = torch.ops.aten.view(sum_92, [384]);  sum_92 = None\0A    detach_224 = torch.ops.aten.detach(view_230);  view_230 = None\0A    detach_225 = torch.ops.aten.detach(detach_224);  detach_224 = None\0A    _reshape_alias_169 = torch.ops.aten._reshape_alias(mm_76, [1, 128, 1536], [196608, 1536, 1]);  mm_76 = None\0A    t_229 = torch.ops.aten.t(t_228);  t_228 = None\0A    detach_226 = torch.ops.aten.detach(t_229);  t_229 = None\0A    detach_227 = torch.ops.aten.detach(detach_226);  detach_226 = None\0A    gelu_backward_6 = torch.ops.aten.gelu_backward(_reshape_alias_169, view_93);  _reshape_alias_169 = view_93 = None\0A    _reshape_alias_170 = torch.ops.aten._reshape_alias(gelu_backward_6, [128, 1536], [1536, 1]);  gelu_backward_6 = None\0A    t_230 = torch.ops.aten.t(t_34);  t_34 = None\0A    mm_78 = torch.ops.aten.mm(_reshape_alias_170, t_230);  t_230 = None\0A    t_231 = torch.ops.aten.t(_reshape_alias_170)\0A    mm_79 = torch.ops.aten.mm(t_231, view_92);  t_231 = view_92 = None\0A    t_232 = torch.ops.aten.t(mm_79);  mm_79 = None\0A    sum_93 = torch.ops.aten.sum(_reshape_alias_170, [0], True);  _reshape_alias_170 = None\0A    view_231 = torch.ops.aten.view(sum_93, [1536]);  sum_93 = None\0A    detach_228 = torch.ops.aten.detach(view_231);  view_231 = None\0A    detach_229 = torch.ops.aten.detach(detach_228);  detach_228 = None\0A    _reshape_alias_171 = torch.ops.aten._reshape_alias(mm_78, [1, 128, 384], [49152, 384, 1]);  mm_78 = None\0A    add_61 = torch.ops.aten.add(mul_89, _reshape_alias_171);  mul_89 = _reshape_alias_171 = None\0A    t_233 = torch.ops.aten.t(t_232);  t_232 = None\0A    detach_230 = torch.ops.aten.detach(t_233);  t_233 = None\0A    detach_231 = torch.ops.aten.detach(detach_230);  detach_230 = None\0A    to_117 = torch.ops.aten.to(add_61, 6)\0A    to_118 = torch.ops.aten.to(add_17, 6)\0A    to_119 = torch.ops.aten.to(getitem_34, 6)\0A    to_120 = torch.ops.aten.to(getitem_35, 6)\0A    to_121 = torch.ops.aten.to(params_119, 6)\0A    to_122 = torch.ops.aten.to(params_118, 6)\0A    sub_39 = torch.ops.aten.sub(add_17, getitem_34);  add_17 = getitem_34 = None\0A    mul_91 = torch.ops.aten.mul(sub_39, getitem_35);  sub_39 = None\0A    mul_92 = torch.ops.aten.mul(add_61, params_119)\0A    mul_93 = torch.ops.aten.mul(mul_92, 384)\0A    sum_94 = torch.ops.aten.sum(mul_92, [2], True)\0A    mul_94 = torch.ops.aten.mul(mul_92, mul_91);  mul_92 = None\0A    sum_95 = torch.ops.aten.sum(mul_94, [2], True);  mul_94 = None\0A    mul_95 = torch.ops.aten.mul(mul_91, sum_95);  sum_95 = None\0A    sub_40 = torch.ops.aten.sub(mul_93, sum_94);  mul_93 = sum_94 = None\0A    sub_41 = torch.ops.aten.sub(sub_40, mul_95);  sub_40 = mul_95 = None\0A    div_31 = torch.ops.aten.div(getitem_35, 384);  getitem_35 = None\0A    mul_96 = torch.ops.aten.mul(div_31, sub_41);  div_31 = sub_41 = None\0A    mul_97 = torch.ops.aten.mul(add_61, mul_91);  mul_91 = None\0A    sum_96 = torch.ops.aten.sum(mul_97, [0, 1]);  mul_97 = None\0A    sum_97 = torch.ops.aten.sum(add_61, [0, 1]);  add_61 = None\0A    to_123 = torch.ops.aten.to(mul_96, 6)\0A    to_124 = torch.ops.aten.to(sum_96, 6)\0A    to_125 = torch.ops.aten.to(sum_97, 6)\0A    detach_232 = torch.ops.aten.detach(sum_96);  sum_96 = None\0A    detach_233 = torch.ops.aten.detach(detach_232);  detach_232 = None\0A    detach_234 = torch.ops.aten.detach(sum_97);  sum_97 = None\0A    detach_235 = torch.ops.aten.detach(detach_234);  detach_234 = None\0A    _reshape_alias_172 = torch.ops.aten._reshape_alias(mul_96, [128, 384], [384, 1])\0A    t_234 = torch.ops.aten.t(t_33);  t_33 = None\0A    mm_80 = torch.ops.aten.mm(_reshape_alias_172, t_234);  t_234 = None\0A    t_235 = torch.ops.aten.t(_reshape_alias_172)\0A    mm_81 = torch.ops.aten.mm(t_235, view_90);  t_235 = view_90 = None\0A    t_236 = torch.ops.aten.t(mm_81);  mm_81 = None\0A    sum_98 = torch.ops.aten.sum(_reshape_alias_172, [0], True);  _reshape_alias_172 = None\0A    view_232 = torch.ops.aten.view(sum_98, [384]);  sum_98 = None\0A    detach_236 = torch.ops.aten.detach(view_232);  view_232 = None\0A    detach_237 = torch.ops.aten.detach(detach_236);  detach_236 = None\0A    _reshape_alias_173 = torch.ops.aten._reshape_alias(mm_80, [1, 128, 384], [49152, 384, 1]);  mm_80 = None\0A    t_237 = torch.ops.aten.t(t_236);  t_236 = None\0A    detach_238 = torch.ops.aten.detach(t_237);  t_237 = None\0A    detach_239 = torch.ops.aten.detach(detach_238);  detach_238 = None\0A    _reshape_alias_174 = torch.ops.aten._reshape_alias(_reshape_alias_173, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_173 = None\0A    permute_72 = torch.ops.aten.permute(_reshape_alias_174, [0, 2, 1, 3]);  _reshape_alias_174 = None\0A    _reshape_alias_175 = torch.ops.aten._reshape_alias(permute_72, [12, 128, 32], [32, 384, 1]);  permute_72 = None\0A    transpose_42 = torch.ops.aten.transpose(_reshape_alias_22, 1, 2);  _reshape_alias_22 = None\0A    bmm_48 = torch.ops.aten.bmm(transpose_42, _reshape_alias_175);  transpose_42 = None\0A    transpose_43 = torch.ops.aten.transpose(_reshape_alias_23, 1, 2);  _reshape_alias_23 = None\0A    bmm_49 = torch.ops.aten.bmm(_reshape_alias_175, transpose_43);  _reshape_alias_175 = transpose_43 = None\0A    _reshape_alias_176 = torch.ops.aten._reshape_alias(bmm_48, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_48 = None\0A    _reshape_alias_177 = torch.ops.aten._reshape_alias(bmm_49, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_49 = None\0A    detach_240 = torch.ops.aten.detach(detach_5);  detach_5 = None\0A    _softmax_backward_data_6 = torch.ops.aten._softmax_backward_data(_reshape_alias_177, detach_240, -1, 6);  _reshape_alias_177 = detach_240 = None\0A    div_32 = torch.ops.aten.div(_softmax_backward_data_6, 5.656854249492381);  _softmax_backward_data_6 = None\0A    _reshape_alias_178 = torch.ops.aten._reshape_alias(div_32, [12, 128, 128], [16384, 128, 1]);  div_32 = None\0A    transpose_44 = torch.ops.aten.transpose(_reshape_alias_20, 1, 2);  _reshape_alias_20 = None\0A    bmm_50 = torch.ops.aten.bmm(transpose_44, _reshape_alias_178);  transpose_44 = None\0A    transpose_45 = torch.ops.aten.transpose(_reshape_alias_21, 1, 2);  _reshape_alias_21 = None\0A    bmm_51 = torch.ops.aten.bmm(_reshape_alias_178, transpose_45);  _reshape_alias_178 = transpose_45 = None\0A    _reshape_alias_179 = torch.ops.aten._reshape_alias(bmm_50, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_50 = None\0A    _reshape_alias_180 = torch.ops.aten._reshape_alias(bmm_51, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_51 = None\0A    transpose_46 = torch.ops.aten.transpose(_reshape_alias_179, -1, -2);  _reshape_alias_179 = None\0A    permute_73 = torch.ops.aten.permute(_reshape_alias_180, [0, 2, 1, 3]);  _reshape_alias_180 = None\0A    clone_24 = torch.ops.aten.clone(permute_73, memory_format = 0);  permute_73 = None\0A    _unsafe_view_36 = torch.ops.aten._unsafe_view(clone_24, [1, 128, 384]);  clone_24 = None\0A    permute_74 = torch.ops.aten.permute(_reshape_alias_176, [0, 2, 1, 3]);  _reshape_alias_176 = None\0A    clone_25 = torch.ops.aten.clone(permute_74, memory_format = 0);  permute_74 = None\0A    _unsafe_view_37 = torch.ops.aten._unsafe_view(clone_25, [1, 128, 384]);  clone_25 = None\0A    _reshape_alias_181 = torch.ops.aten._reshape_alias(_unsafe_view_37, [128, 384], [384, 1]);  _unsafe_view_37 = None\0A    t_238 = torch.ops.aten.t(t_32);  t_32 = None\0A    mm_82 = torch.ops.aten.mm(_reshape_alias_181, t_238);  t_238 = None\0A    t_239 = torch.ops.aten.t(_reshape_alias_181)\0A    mm_83 = torch.ops.aten.mm(t_239, view_85);  t_239 = view_85 = None\0A    t_240 = torch.ops.aten.t(mm_83);  mm_83 = None\0A    sum_99 = torch.ops.aten.sum(_reshape_alias_181, [0], True);  _reshape_alias_181 = None\0A    view_233 = torch.ops.aten.view(sum_99, [384]);  sum_99 = None\0A    detach_241 = torch.ops.aten.detach(view_233);  view_233 = None\0A    detach_242 = torch.ops.aten.detach(detach_241);  detach_241 = None\0A    _reshape_alias_182 = torch.ops.aten._reshape_alias(mm_82, [1, 128, 384], [49152, 384, 1]);  mm_82 = None\0A    add_62 = torch.ops.aten.add(mul_96, _reshape_alias_182);  mul_96 = _reshape_alias_182 = None\0A    t_241 = torch.ops.aten.t(t_240);  t_240 = None\0A    detach_243 = torch.ops.aten.detach(t_241);  t_241 = None\0A    detach_244 = torch.ops.aten.detach(detach_243);  detach_243 = None\0A    permute_75 = torch.ops.aten.permute(transpose_46, [0, 2, 1, 3]);  transpose_46 = None\0A    _reshape_alias_183 = torch.ops.aten._reshape_alias(permute_75, [1, 128, 384], [128, 1, 128]);  permute_75 = None\0A    _reshape_alias_184 = torch.ops.aten._reshape_alias(_reshape_alias_183, [128, 384], [1, 128]);  _reshape_alias_183 = None\0A    t_242 = torch.ops.aten.t(t_31);  t_31 = None\0A    mm_84 = torch.ops.aten.mm(_reshape_alias_184, t_242);  t_242 = None\0A    t_243 = torch.ops.aten.t(_reshape_alias_184)\0A    mm_85 = torch.ops.aten.mm(t_243, view_82);  t_243 = view_82 = None\0A    t_244 = torch.ops.aten.t(mm_85);  mm_85 = None\0A    sum_100 = torch.ops.aten.sum(_reshape_alias_184, [0], True);  _reshape_alias_184 = None\0A    view_234 = torch.ops.aten.view(sum_100, [384]);  sum_100 = None\0A    detach_245 = torch.ops.aten.detach(view_234);  view_234 = None\0A    detach_246 = torch.ops.aten.detach(detach_245);  detach_245 = None\0A    _reshape_alias_185 = torch.ops.aten._reshape_alias(mm_84, [1, 128, 384], [49152, 384, 1]);  mm_84 = None\0A    add_63 = torch.ops.aten.add(add_62, _reshape_alias_185);  add_62 = _reshape_alias_185 = None\0A    t_245 = torch.ops.aten.t(t_244);  t_244 = None\0A    detach_247 = torch.ops.aten.detach(t_245);  t_245 = None\0A    detach_248 = torch.ops.aten.detach(detach_247);  detach_247 = None\0A    _reshape_alias_186 = torch.ops.aten._reshape_alias(_unsafe_view_36, [128, 384], [384, 1]);  _unsafe_view_36 = None\0A    t_246 = torch.ops.aten.t(t_30);  t_30 = None\0A    mm_86 = torch.ops.aten.mm(_reshape_alias_186, t_246);  t_246 = None\0A    t_247 = torch.ops.aten.t(_reshape_alias_186)\0A    mm_87 = torch.ops.aten.mm(t_247, view_80);  t_247 = view_80 = None\0A    t_248 = torch.ops.aten.t(mm_87);  mm_87 = None\0A    sum_101 = torch.ops.aten.sum(_reshape_alias_186, [0], True);  _reshape_alias_186 = None\0A    view_235 = torch.ops.aten.view(sum_101, [384]);  sum_101 = None\0A    detach_249 = torch.ops.aten.detach(view_235);  view_235 = None\0A    detach_250 = torch.ops.aten.detach(detach_249);  detach_249 = None\0A    _reshape_alias_187 = torch.ops.aten._reshape_alias(mm_86, [1, 128, 384], [49152, 384, 1]);  mm_86 = None\0A    add_64 = torch.ops.aten.add(add_63, _reshape_alias_187);  add_63 = _reshape_alias_187 = None\0A    t_249 = torch.ops.aten.t(t_248);  t_248 = None\0A    detach_251 = torch.ops.aten.detach(t_249);  t_249 = None\0A    detach_252 = torch.ops.aten.detach(detach_251);  detach_251 = None\0A    to_126 = torch.ops.aten.to(add_64, 6)\0A    to_127 = torch.ops.aten.to(add_15, 6)\0A    to_128 = torch.ops.aten.to(getitem_31, 6)\0A    to_129 = torch.ops.aten.to(getitem_32, 6)\0A    to_130 = torch.ops.aten.to(params_115, 6)\0A    to_131 = torch.ops.aten.to(params_114, 6)\0A    sub_42 = torch.ops.aten.sub(add_15, getitem_31);  add_15 = getitem_31 = None\0A    mul_98 = torch.ops.aten.mul(sub_42, getitem_32);  sub_42 = None\0A    mul_99 = torch.ops.aten.mul(add_64, params_115)\0A    mul_100 = torch.ops.aten.mul(mul_99, 384)\0A    sum_102 = torch.ops.aten.sum(mul_99, [2], True)\0A    mul_101 = torch.ops.aten.mul(mul_99, mul_98);  mul_99 = None\0A    sum_103 = torch.ops.aten.sum(mul_101, [2], True);  mul_101 = None\0A    mul_102 = torch.ops.aten.mul(mul_98, sum_103);  sum_103 = None\0A    sub_43 = torch.ops.aten.sub(mul_100, sum_102);  mul_100 = sum_102 = None\0A    sub_44 = torch.ops.aten.sub(sub_43, mul_102);  sub_43 = mul_102 = None\0A    div_33 = torch.ops.aten.div(getitem_32, 384);  getitem_32 = None\0A    mul_103 = torch.ops.aten.mul(div_33, sub_44);  div_33 = sub_44 = None\0A    mul_104 = torch.ops.aten.mul(add_64, mul_98);  mul_98 = None\0A    sum_104 = torch.ops.aten.sum(mul_104, [0, 1]);  mul_104 = None\0A    sum_105 = torch.ops.aten.sum(add_64, [0, 1]);  add_64 = None\0A    to_132 = torch.ops.aten.to(mul_103, 6)\0A    to_133 = torch.ops.aten.to(sum_104, 6)\0A    to_134 = torch.ops.aten.to(sum_105, 6)\0A    detach_253 = torch.ops.aten.detach(sum_104);  sum_104 = None\0A    detach_254 = torch.ops.aten.detach(detach_253);  detach_253 = None\0A    detach_255 = torch.ops.aten.detach(sum_105);  sum_105 = None\0A    detach_256 = torch.ops.aten.detach(detach_255);  detach_255 = None\0A    _reshape_alias_188 = torch.ops.aten._reshape_alias(mul_103, [128, 384], [384, 1])\0A    t_250 = torch.ops.aten.t(t_29);  t_29 = None\0A    mm_88 = torch.ops.aten.mm(_reshape_alias_188, t_250);  t_250 = None\0A    t_251 = torch.ops.aten.t(_reshape_alias_188)\0A    mm_89 = torch.ops.aten.mm(t_251, view_78);  t_251 = view_78 = None\0A    t_252 = torch.ops.aten.t(mm_89);  mm_89 = None\0A    sum_106 = torch.ops.aten.sum(_reshape_alias_188, [0], True);  _reshape_alias_188 = None\0A    view_236 = torch.ops.aten.view(sum_106, [384]);  sum_106 = None\0A    detach_257 = torch.ops.aten.detach(view_236);  view_236 = None\0A    detach_258 = torch.ops.aten.detach(detach_257);  detach_257 = None\0A    _reshape_alias_189 = torch.ops.aten._reshape_alias(mm_88, [1, 128, 1536], [196608, 1536, 1]);  mm_88 = None\0A    t_253 = torch.ops.aten.t(t_252);  t_252 = None\0A    detach_259 = torch.ops.aten.detach(t_253);  t_253 = None\0A    detach_260 = torch.ops.aten.detach(detach_259);  detach_259 = None\0A    gelu_backward_7 = torch.ops.aten.gelu_backward(_reshape_alias_189, view_77);  _reshape_alias_189 = view_77 = None\0A    _reshape_alias_190 = torch.ops.aten._reshape_alias(gelu_backward_7, [128, 1536], [1536, 1]);  gelu_backward_7 = None\0A    t_254 = torch.ops.aten.t(t_28);  t_28 = None\0A    mm_90 = torch.ops.aten.mm(_reshape_alias_190, t_254);  t_254 = None\0A    t_255 = torch.ops.aten.t(_reshape_alias_190)\0A    mm_91 = torch.ops.aten.mm(t_255, view_76);  t_255 = view_76 = None\0A    t_256 = torch.ops.aten.t(mm_91);  mm_91 = None\0A    sum_107 = torch.ops.aten.sum(_reshape_alias_190, [0], True);  _reshape_alias_190 = None\0A    view_237 = torch.ops.aten.view(sum_107, [1536]);  sum_107 = None\0A    detach_261 = torch.ops.aten.detach(view_237);  view_237 = None\0A    detach_262 = torch.ops.aten.detach(detach_261);  detach_261 = None\0A    _reshape_alias_191 = torch.ops.aten._reshape_alias(mm_90, [1, 128, 384], [49152, 384, 1]);  mm_90 = None\0A    add_65 = torch.ops.aten.add(mul_103, _reshape_alias_191);  mul_103 = _reshape_alias_191 = None\0A    t_257 = torch.ops.aten.t(t_256);  t_256 = None\0A    detach_263 = torch.ops.aten.detach(t_257);  t_257 = None\0A    detach_264 = torch.ops.aten.detach(detach_263);  detach_263 = None\0A    to_135 = torch.ops.aten.to(add_65, 6)\0A    to_136 = torch.ops.aten.to(add_14, 6)\0A    to_137 = torch.ops.aten.to(getitem_28, 6)\0A    to_138 = torch.ops.aten.to(getitem_29, 6)\0A    to_139 = torch.ops.aten.to(params_103, 6)\0A    to_140 = torch.ops.aten.to(params_102, 6)\0A    sub_45 = torch.ops.aten.sub(add_14, getitem_28);  add_14 = getitem_28 = None\0A    mul_105 = torch.ops.aten.mul(sub_45, getitem_29);  sub_45 = None\0A    mul_106 = torch.ops.aten.mul(add_65, params_103)\0A    mul_107 = torch.ops.aten.mul(mul_106, 384)\0A    sum_108 = torch.ops.aten.sum(mul_106, [2], True)\0A    mul_108 = torch.ops.aten.mul(mul_106, mul_105);  mul_106 = None\0A    sum_109 = torch.ops.aten.sum(mul_108, [2], True);  mul_108 = None\0A    mul_109 = torch.ops.aten.mul(mul_105, sum_109);  sum_109 = None\0A    sub_46 = torch.ops.aten.sub(mul_107, sum_108);  mul_107 = sum_108 = None\0A    sub_47 = torch.ops.aten.sub(sub_46, mul_109);  sub_46 = mul_109 = None\0A    div_34 = torch.ops.aten.div(getitem_29, 384);  getitem_29 = None\0A    mul_110 = torch.ops.aten.mul(div_34, sub_47);  div_34 = sub_47 = None\0A    mul_111 = torch.ops.aten.mul(add_65, mul_105);  mul_105 = None\0A    sum_110 = torch.ops.aten.sum(mul_111, [0, 1]);  mul_111 = None\0A    sum_111 = torch.ops.aten.sum(add_65, [0, 1]);  add_65 = None\0A    to_141 = torch.ops.aten.to(mul_110, 6)\0A    to_142 = torch.ops.aten.to(sum_110, 6)\0A    to_143 = torch.ops.aten.to(sum_111, 6)\0A    detach_265 = torch.ops.aten.detach(sum_110);  sum_110 = None\0A    detach_266 = torch.ops.aten.detach(detach_265);  detach_265 = None\0A    detach_267 = torch.ops.aten.detach(sum_111);  sum_111 = None\0A    detach_268 = torch.ops.aten.detach(detach_267);  detach_267 = None\0A    _reshape_alias_192 = torch.ops.aten._reshape_alias(mul_110, [128, 384], [384, 1])\0A    t_258 = torch.ops.aten.t(t_27);  t_27 = None\0A    mm_92 = torch.ops.aten.mm(_reshape_alias_192, t_258);  t_258 = None\0A    t_259 = torch.ops.aten.t(_reshape_alias_192)\0A    mm_93 = torch.ops.aten.mm(t_259, view_74);  t_259 = view_74 = None\0A    t_260 = torch.ops.aten.t(mm_93);  mm_93 = None\0A    sum_112 = torch.ops.aten.sum(_reshape_alias_192, [0], True);  _reshape_alias_192 = None\0A    view_238 = torch.ops.aten.view(sum_112, [384]);  sum_112 = None\0A    detach_269 = torch.ops.aten.detach(view_238);  view_238 = None\0A    detach_270 = torch.ops.aten.detach(detach_269);  detach_269 = None\0A    _reshape_alias_193 = torch.ops.aten._reshape_alias(mm_92, [1, 128, 384], [49152, 384, 1]);  mm_92 = None\0A    t_261 = torch.ops.aten.t(t_260);  t_260 = None\0A    detach_271 = torch.ops.aten.detach(t_261);  t_261 = None\0A    detach_272 = torch.ops.aten.detach(detach_271);  detach_271 = None\0A    _reshape_alias_194 = torch.ops.aten._reshape_alias(_reshape_alias_193, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_193 = None\0A    permute_76 = torch.ops.aten.permute(_reshape_alias_194, [0, 2, 1, 3]);  _reshape_alias_194 = None\0A    _reshape_alias_195 = torch.ops.aten._reshape_alias(permute_76, [12, 128, 32], [32, 384, 1]);  permute_76 = None\0A    transpose_47 = torch.ops.aten.transpose(_reshape_alias_18, 1, 2);  _reshape_alias_18 = None\0A    bmm_52 = torch.ops.aten.bmm(transpose_47, _reshape_alias_195);  transpose_47 = None\0A    transpose_48 = torch.ops.aten.transpose(_reshape_alias_19, 1, 2);  _reshape_alias_19 = None\0A    bmm_53 = torch.ops.aten.bmm(_reshape_alias_195, transpose_48);  _reshape_alias_195 = transpose_48 = None\0A    _reshape_alias_196 = torch.ops.aten._reshape_alias(bmm_52, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_52 = None\0A    _reshape_alias_197 = torch.ops.aten._reshape_alias(bmm_53, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_53 = None\0A    detach_273 = torch.ops.aten.detach(detach_4);  detach_4 = None\0A    _softmax_backward_data_7 = torch.ops.aten._softmax_backward_data(_reshape_alias_197, detach_273, -1, 6);  _reshape_alias_197 = detach_273 = None\0A    div_35 = torch.ops.aten.div(_softmax_backward_data_7, 5.656854249492381);  _softmax_backward_data_7 = None\0A    _reshape_alias_198 = torch.ops.aten._reshape_alias(div_35, [12, 128, 128], [16384, 128, 1]);  div_35 = None\0A    transpose_49 = torch.ops.aten.transpose(_reshape_alias_16, 1, 2);  _reshape_alias_16 = None\0A    bmm_54 = torch.ops.aten.bmm(transpose_49, _reshape_alias_198);  transpose_49 = None\0A    transpose_50 = torch.ops.aten.transpose(_reshape_alias_17, 1, 2);  _reshape_alias_17 = None\0A    bmm_55 = torch.ops.aten.bmm(_reshape_alias_198, transpose_50);  _reshape_alias_198 = transpose_50 = None\0A    _reshape_alias_199 = torch.ops.aten._reshape_alias(bmm_54, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_54 = None\0A    _reshape_alias_200 = torch.ops.aten._reshape_alias(bmm_55, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_55 = None\0A    transpose_51 = torch.ops.aten.transpose(_reshape_alias_199, -1, -2);  _reshape_alias_199 = None\0A    permute_77 = torch.ops.aten.permute(_reshape_alias_200, [0, 2, 1, 3]);  _reshape_alias_200 = None\0A    clone_26 = torch.ops.aten.clone(permute_77, memory_format = 0);  permute_77 = None\0A    _unsafe_view_38 = torch.ops.aten._unsafe_view(clone_26, [1, 128, 384]);  clone_26 = None\0A    permute_78 = torch.ops.aten.permute(_reshape_alias_196, [0, 2, 1, 3]);  _reshape_alias_196 = None\0A    clone_27 = torch.ops.aten.clone(permute_78, memory_format = 0);  permute_78 = None\0A    _unsafe_view_39 = torch.ops.aten._unsafe_view(clone_27, [1, 128, 384]);  clone_27 = None\0A    _reshape_alias_201 = torch.ops.aten._reshape_alias(_unsafe_view_39, [128, 384], [384, 1]);  _unsafe_view_39 = None\0A    t_262 = torch.ops.aten.t(t_26);  t_26 = None\0A    mm_94 = torch.ops.aten.mm(_reshape_alias_201, t_262);  t_262 = None\0A    t_263 = torch.ops.aten.t(_reshape_alias_201)\0A    mm_95 = torch.ops.aten.mm(t_263, view_69);  t_263 = view_69 = None\0A    t_264 = torch.ops.aten.t(mm_95);  mm_95 = None\0A    sum_113 = torch.ops.aten.sum(_reshape_alias_201, [0], True);  _reshape_alias_201 = None\0A    view_239 = torch.ops.aten.view(sum_113, [384]);  sum_113 = None\0A    detach_274 = torch.ops.aten.detach(view_239);  view_239 = None\0A    detach_275 = torch.ops.aten.detach(detach_274);  detach_274 = None\0A    _reshape_alias_202 = torch.ops.aten._reshape_alias(mm_94, [1, 128, 384], [49152, 384, 1]);  mm_94 = None\0A    add_66 = torch.ops.aten.add(mul_110, _reshape_alias_202);  mul_110 = _reshape_alias_202 = None\0A    t_265 = torch.ops.aten.t(t_264);  t_264 = None\0A    detach_276 = torch.ops.aten.detach(t_265);  t_265 = None\0A    detach_277 = torch.ops.aten.detach(detach_276);  detach_276 = None\0A    permute_79 = torch.ops.aten.permute(transpose_51, [0, 2, 1, 3]);  transpose_51 = None\0A    _reshape_alias_203 = torch.ops.aten._reshape_alias(permute_79, [1, 128, 384], [128, 1, 128]);  permute_79 = None\0A    _reshape_alias_204 = torch.ops.aten._reshape_alias(_reshape_alias_203, [128, 384], [1, 128]);  _reshape_alias_203 = None\0A    t_266 = torch.ops.aten.t(t_25);  t_25 = None\0A    mm_96 = torch.ops.aten.mm(_reshape_alias_204, t_266);  t_266 = None\0A    t_267 = torch.ops.aten.t(_reshape_alias_204)\0A    mm_97 = torch.ops.aten.mm(t_267, view_66);  t_267 = view_66 = None\0A    t_268 = torch.ops.aten.t(mm_97);  mm_97 = None\0A    sum_114 = torch.ops.aten.sum(_reshape_alias_204, [0], True);  _reshape_alias_204 = None\0A    view_240 = torch.ops.aten.view(sum_114, [384]);  sum_114 = None\0A    detach_278 = torch.ops.aten.detach(view_240);  view_240 = None\0A    detach_279 = torch.ops.aten.detach(detach_278);  detach_278 = None\0A    _reshape_alias_205 = torch.ops.aten._reshape_alias(mm_96, [1, 128, 384], [49152, 384, 1]);  mm_96 = None\0A    add_67 = torch.ops.aten.add(add_66, _reshape_alias_205);  add_66 = _reshape_alias_205 = None\0A    t_269 = torch.ops.aten.t(t_268);  t_268 = None\0A    detach_280 = torch.ops.aten.detach(t_269);  t_269 = None\0A    detach_281 = torch.ops.aten.detach(detach_280);  detach_280 = None\0A    _reshape_alias_206 = torch.ops.aten._reshape_alias(_unsafe_view_38, [128, 384], [384, 1]);  _unsafe_view_38 = None\0A    t_270 = torch.ops.aten.t(t_24);  t_24 = None\0A    mm_98 = torch.ops.aten.mm(_reshape_alias_206, t_270);  t_270 = None\0A    t_271 = torch.ops.aten.t(_reshape_alias_206)\0A    mm_99 = torch.ops.aten.mm(t_271, view_64);  t_271 = view_64 = None\0A    t_272 = torch.ops.aten.t(mm_99);  mm_99 = None\0A    sum_115 = torch.ops.aten.sum(_reshape_alias_206, [0], True);  _reshape_alias_206 = None\0A    view_241 = torch.ops.aten.view(sum_115, [384]);  sum_115 = None\0A    detach_282 = torch.ops.aten.detach(view_241);  view_241 = None\0A    detach_283 = torch.ops.aten.detach(detach_282);  detach_282 = None\0A    _reshape_alias_207 = torch.ops.aten._reshape_alias(mm_98, [1, 128, 384], [49152, 384, 1]);  mm_98 = None\0A    add_68 = torch.ops.aten.add(add_67, _reshape_alias_207);  add_67 = _reshape_alias_207 = None\0A    t_273 = torch.ops.aten.t(t_272);  t_272 = None\0A    detach_284 = torch.ops.aten.detach(t_273);  t_273 = None\0A    detach_285 = torch.ops.aten.detach(detach_284);  detach_284 = None\0A    to_144 = torch.ops.aten.to(add_68, 6)\0A    to_145 = torch.ops.aten.to(add_12, 6)\0A    to_146 = torch.ops.aten.to(getitem_25, 6)\0A    to_147 = torch.ops.aten.to(getitem_26, 6)\0A    to_148 = torch.ops.aten.to(params_99, 6)\0A    to_149 = torch.ops.aten.to(params_98, 6)\0A    sub_48 = torch.ops.aten.sub(add_12, getitem_25);  add_12 = getitem_25 = None\0A    mul_112 = torch.ops.aten.mul(sub_48, getitem_26);  sub_48 = None\0A    mul_113 = torch.ops.aten.mul(add_68, params_99)\0A    mul_114 = torch.ops.aten.mul(mul_113, 384)\0A    sum_116 = torch.ops.aten.sum(mul_113, [2], True)\0A    mul_115 = torch.ops.aten.mul(mul_113, mul_112);  mul_113 = None\0A    sum_117 = torch.ops.aten.sum(mul_115, [2], True);  mul_115 = None\0A    mul_116 = torch.ops.aten.mul(mul_112, sum_117);  sum_117 = None\0A    sub_49 = torch.ops.aten.sub(mul_114, sum_116);  mul_114 = sum_116 = None\0A    sub_50 = torch.ops.aten.sub(sub_49, mul_116);  sub_49 = mul_116 = None\0A    div_36 = torch.ops.aten.div(getitem_26, 384);  getitem_26 = None\0A    mul_117 = torch.ops.aten.mul(div_36, sub_50);  div_36 = sub_50 = None\0A    mul_118 = torch.ops.aten.mul(add_68, mul_112);  mul_112 = None\0A    sum_118 = torch.ops.aten.sum(mul_118, [0, 1]);  mul_118 = None\0A    sum_119 = torch.ops.aten.sum(add_68, [0, 1]);  add_68 = None\0A    to_150 = torch.ops.aten.to(mul_117, 6)\0A    to_151 = torch.ops.aten.to(sum_118, 6)\0A    to_152 = torch.ops.aten.to(sum_119, 6)\0A    detach_286 = torch.ops.aten.detach(sum_118);  sum_118 = None\0A    detach_287 = torch.ops.aten.detach(detach_286);  detach_286 = None\0A    detach_288 = torch.ops.aten.detach(sum_119);  sum_119 = None\0A    detach_289 = torch.ops.aten.detach(detach_288);  detach_288 = None\0A    _reshape_alias_208 = torch.ops.aten._reshape_alias(mul_117, [128, 384], [384, 1])\0A    t_274 = torch.ops.aten.t(t_23);  t_23 = None\0A    mm_100 = torch.ops.aten.mm(_reshape_alias_208, t_274);  t_274 = None\0A    t_275 = torch.ops.aten.t(_reshape_alias_208)\0A    mm_101 = torch.ops.aten.mm(t_275, view_62);  t_275 = view_62 = None\0A    t_276 = torch.ops.aten.t(mm_101);  mm_101 = None\0A    sum_120 = torch.ops.aten.sum(_reshape_alias_208, [0], True);  _reshape_alias_208 = None\0A    view_242 = torch.ops.aten.view(sum_120, [384]);  sum_120 = None\0A    detach_290 = torch.ops.aten.detach(view_242);  view_242 = None\0A    detach_291 = torch.ops.aten.detach(detach_290);  detach_290 = None\0A    _reshape_alias_209 = torch.ops.aten._reshape_alias(mm_100, [1, 128, 1536], [196608, 1536, 1]);  mm_100 = None\0A    t_277 = torch.ops.aten.t(t_276);  t_276 = None\0A    detach_292 = torch.ops.aten.detach(t_277);  t_277 = None\0A    detach_293 = torch.ops.aten.detach(detach_292);  detach_292 = None\0A    gelu_backward_8 = torch.ops.aten.gelu_backward(_reshape_alias_209, view_61);  _reshape_alias_209 = view_61 = None\0A    _reshape_alias_210 = torch.ops.aten._reshape_alias(gelu_backward_8, [128, 1536], [1536, 1]);  gelu_backward_8 = None\0A    t_278 = torch.ops.aten.t(t_22);  t_22 = None\0A    mm_102 = torch.ops.aten.mm(_reshape_alias_210, t_278);  t_278 = None\0A    t_279 = torch.ops.aten.t(_reshape_alias_210)\0A    mm_103 = torch.ops.aten.mm(t_279, view_60);  t_279 = view_60 = None\0A    t_280 = torch.ops.aten.t(mm_103);  mm_103 = None\0A    sum_121 = torch.ops.aten.sum(_reshape_alias_210, [0], True);  _reshape_alias_210 = None\0A    view_243 = torch.ops.aten.view(sum_121, [1536]);  sum_121 = None\0A    detach_294 = torch.ops.aten.detach(view_243);  view_243 = None\0A    detach_295 = torch.ops.aten.detach(detach_294);  detach_294 = None\0A    _reshape_alias_211 = torch.ops.aten._reshape_alias(mm_102, [1, 128, 384], [49152, 384, 1]);  mm_102 = None\0A    add_69 = torch.ops.aten.add(mul_117, _reshape_alias_211);  mul_117 = _reshape_alias_211 = None\0A    t_281 = torch.ops.aten.t(t_280);  t_280 = None\0A    detach_296 = torch.ops.aten.detach(t_281);  t_281 = None\0A    detach_297 = torch.ops.aten.detach(detach_296);  detach_296 = None\0A    to_153 = torch.ops.aten.to(add_69, 6)\0A    to_154 = torch.ops.aten.to(add_11, 6)\0A    to_155 = torch.ops.aten.to(getitem_22, 6)\0A    to_156 = torch.ops.aten.to(getitem_23, 6)\0A    to_157 = torch.ops.aten.to(params_87, 6)\0A    to_158 = torch.ops.aten.to(params_86, 6)\0A    sub_51 = torch.ops.aten.sub(add_11, getitem_22);  add_11 = getitem_22 = None\0A    mul_119 = torch.ops.aten.mul(sub_51, getitem_23);  sub_51 = None\0A    mul_120 = torch.ops.aten.mul(add_69, params_87)\0A    mul_121 = torch.ops.aten.mul(mul_120, 384)\0A    sum_122 = torch.ops.aten.sum(mul_120, [2], True)\0A    mul_122 = torch.ops.aten.mul(mul_120, mul_119);  mul_120 = None\0A    sum_123 = torch.ops.aten.sum(mul_122, [2], True);  mul_122 = None\0A    mul_123 = torch.ops.aten.mul(mul_119, sum_123);  sum_123 = None\0A    sub_52 = torch.ops.aten.sub(mul_121, sum_122);  mul_121 = sum_122 = None\0A    sub_53 = torch.ops.aten.sub(sub_52, mul_123);  sub_52 = mul_123 = None\0A    div_37 = torch.ops.aten.div(getitem_23, 384);  getitem_23 = None\0A    mul_124 = torch.ops.aten.mul(div_37, sub_53);  div_37 = sub_53 = None\0A    mul_125 = torch.ops.aten.mul(add_69, mul_119);  mul_119 = None\0A    sum_124 = torch.ops.aten.sum(mul_125, [0, 1]);  mul_125 = None\0A    sum_125 = torch.ops.aten.sum(add_69, [0, 1]);  add_69 = None\0A    to_159 = torch.ops.aten.to(mul_124, 6)\0A    to_160 = torch.ops.aten.to(sum_124, 6)\0A    to_161 = torch.ops.aten.to(sum_125, 6)\0A    detach_298 = torch.ops.aten.detach(sum_124);  sum_124 = None\0A    detach_299 = torch.ops.aten.detach(detach_298);  detach_298 = None\0A    detach_300 = torch.ops.aten.detach(sum_125);  sum_125 = None\0A    detach_301 = torch.ops.aten.detach(detach_300);  detach_300 = None\0A    _reshape_alias_212 = torch.ops.aten._reshape_alias(mul_124, [128, 384], [384, 1])\0A    t_282 = torch.ops.aten.t(t_21);  t_21 = None\0A    mm_104 = torch.ops.aten.mm(_reshape_alias_212, t_282);  t_282 = None\0A    t_283 = torch.ops.aten.t(_reshape_alias_212)\0A    mm_105 = torch.ops.aten.mm(t_283, view_58);  t_283 = view_58 = None\0A    t_284 = torch.ops.aten.t(mm_105);  mm_105 = None\0A    sum_126 = torch.ops.aten.sum(_reshape_alias_212, [0], True);  _reshape_alias_212 = None\0A    view_244 = torch.ops.aten.view(sum_126, [384]);  sum_126 = None\0A    detach_302 = torch.ops.aten.detach(view_244);  view_244 = None\0A    detach_303 = torch.ops.aten.detach(detach_302);  detach_302 = None\0A    _reshape_alias_213 = torch.ops.aten._reshape_alias(mm_104, [1, 128, 384], [49152, 384, 1]);  mm_104 = None\0A    t_285 = torch.ops.aten.t(t_284);  t_284 = None\0A    detach_304 = torch.ops.aten.detach(t_285);  t_285 = None\0A    detach_305 = torch.ops.aten.detach(detach_304);  detach_304 = None\0A    _reshape_alias_214 = torch.ops.aten._reshape_alias(_reshape_alias_213, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_213 = None\0A    permute_80 = torch.ops.aten.permute(_reshape_alias_214, [0, 2, 1, 3]);  _reshape_alias_214 = None\0A    _reshape_alias_215 = torch.ops.aten._reshape_alias(permute_80, [12, 128, 32], [32, 384, 1]);  permute_80 = None\0A    transpose_52 = torch.ops.aten.transpose(_reshape_alias_14, 1, 2);  _reshape_alias_14 = None\0A    bmm_56 = torch.ops.aten.bmm(transpose_52, _reshape_alias_215);  transpose_52 = None\0A    transpose_53 = torch.ops.aten.transpose(_reshape_alias_15, 1, 2);  _reshape_alias_15 = None\0A    bmm_57 = torch.ops.aten.bmm(_reshape_alias_215, transpose_53);  _reshape_alias_215 = transpose_53 = None\0A    _reshape_alias_216 = torch.ops.aten._reshape_alias(bmm_56, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_56 = None\0A    _reshape_alias_217 = torch.ops.aten._reshape_alias(bmm_57, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_57 = None\0A    detach_306 = torch.ops.aten.detach(detach_3);  detach_3 = None\0A    _softmax_backward_data_8 = torch.ops.aten._softmax_backward_data(_reshape_alias_217, detach_306, -1, 6);  _reshape_alias_217 = detach_306 = None\0A    div_38 = torch.ops.aten.div(_softmax_backward_data_8, 5.656854249492381);  _softmax_backward_data_8 = None\0A    _reshape_alias_218 = torch.ops.aten._reshape_alias(div_38, [12, 128, 128], [16384, 128, 1]);  div_38 = None\0A    transpose_54 = torch.ops.aten.transpose(_reshape_alias_12, 1, 2);  _reshape_alias_12 = None\0A    bmm_58 = torch.ops.aten.bmm(transpose_54, _reshape_alias_218);  transpose_54 = None\0A    transpose_55 = torch.ops.aten.transpose(_reshape_alias_13, 1, 2);  _reshape_alias_13 = None\0A    bmm_59 = torch.ops.aten.bmm(_reshape_alias_218, transpose_55);  _reshape_alias_218 = transpose_55 = None\0A    _reshape_alias_219 = torch.ops.aten._reshape_alias(bmm_58, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_58 = None\0A    _reshape_alias_220 = torch.ops.aten._reshape_alias(bmm_59, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_59 = None\0A    transpose_56 = torch.ops.aten.transpose(_reshape_alias_219, -1, -2);  _reshape_alias_219 = None\0A    permute_81 = torch.ops.aten.permute(_reshape_alias_220, [0, 2, 1, 3]);  _reshape_alias_220 = None\0A    clone_28 = torch.ops.aten.clone(permute_81, memory_format = 0);  permute_81 = None\0A    _unsafe_view_40 = torch.ops.aten._unsafe_view(clone_28, [1, 128, 384]);  clone_28 = None\0A    permute_82 = torch.ops.aten.permute(_reshape_alias_216, [0, 2, 1, 3]);  _reshape_alias_216 = None\0A    clone_29 = torch.ops.aten.clone(permute_82, memory_format = 0);  permute_82 = None\0A    _unsafe_view_41 = torch.ops.aten._unsafe_view(clone_29, [1, 128, 384]);  clone_29 = None\0A    _reshape_alias_221 = torch.ops.aten._reshape_alias(_unsafe_view_41, [128, 384], [384, 1]);  _unsafe_view_41 = None\0A    t_286 = torch.ops.aten.t(t_20);  t_20 = None\0A    mm_106 = torch.ops.aten.mm(_reshape_alias_221, t_286);  t_286 = None\0A    t_287 = torch.ops.aten.t(_reshape_alias_221)\0A    mm_107 = torch.ops.aten.mm(t_287, view_53);  t_287 = view_53 = None\0A    t_288 = torch.ops.aten.t(mm_107);  mm_107 = None\0A    sum_127 = torch.ops.aten.sum(_reshape_alias_221, [0], True);  _reshape_alias_221 = None\0A    view_245 = torch.ops.aten.view(sum_127, [384]);  sum_127 = None\0A    detach_307 = torch.ops.aten.detach(view_245);  view_245 = None\0A    detach_308 = torch.ops.aten.detach(detach_307);  detach_307 = None\0A    _reshape_alias_222 = torch.ops.aten._reshape_alias(mm_106, [1, 128, 384], [49152, 384, 1]);  mm_106 = None\0A    add_70 = torch.ops.aten.add(mul_124, _reshape_alias_222);  mul_124 = _reshape_alias_222 = None\0A    t_289 = torch.ops.aten.t(t_288);  t_288 = None\0A    detach_309 = torch.ops.aten.detach(t_289);  t_289 = None\0A    detach_310 = torch.ops.aten.detach(detach_309);  detach_309 = None\0A    permute_83 = torch.ops.aten.permute(transpose_56, [0, 2, 1, 3]);  transpose_56 = None\0A    _reshape_alias_223 = torch.ops.aten._reshape_alias(permute_83, [1, 128, 384], [128, 1, 128]);  permute_83 = None\0A    _reshape_alias_224 = torch.ops.aten._reshape_alias(_reshape_alias_223, [128, 384], [1, 128]);  _reshape_alias_223 = None\0A    t_290 = torch.ops.aten.t(t_19);  t_19 = None\0A    mm_108 = torch.ops.aten.mm(_reshape_alias_224, t_290);  t_290 = None\0A    t_291 = torch.ops.aten.t(_reshape_alias_224)\0A    mm_109 = torch.ops.aten.mm(t_291, view_50);  t_291 = view_50 = None\0A    t_292 = torch.ops.aten.t(mm_109);  mm_109 = None\0A    sum_128 = torch.ops.aten.sum(_reshape_alias_224, [0], True);  _reshape_alias_224 = None\0A    view_246 = torch.ops.aten.view(sum_128, [384]);  sum_128 = None\0A    detach_311 = torch.ops.aten.detach(view_246);  view_246 = None\0A    detach_312 = torch.ops.aten.detach(detach_311);  detach_311 = None\0A    _reshape_alias_225 = torch.ops.aten._reshape_alias(mm_108, [1, 128, 384], [49152, 384, 1]);  mm_108 = None\0A    add_71 = torch.ops.aten.add(add_70, _reshape_alias_225);  add_70 = _reshape_alias_225 = None\0A    t_293 = torch.ops.aten.t(t_292);  t_292 = None\0A    detach_313 = torch.ops.aten.detach(t_293);  t_293 = None\0A    detach_314 = torch.ops.aten.detach(detach_313);  detach_313 = None\0A    _reshape_alias_226 = torch.ops.aten._reshape_alias(_unsafe_view_40, [128, 384], [384, 1]);  _unsafe_view_40 = None\0A    t_294 = torch.ops.aten.t(t_18);  t_18 = None\0A    mm_110 = torch.ops.aten.mm(_reshape_alias_226, t_294);  t_294 = None\0A    t_295 = torch.ops.aten.t(_reshape_alias_226)\0A    mm_111 = torch.ops.aten.mm(t_295, view_48);  t_295 = view_48 = None\0A    t_296 = torch.ops.aten.t(mm_111);  mm_111 = None\0A    sum_129 = torch.ops.aten.sum(_reshape_alias_226, [0], True);  _reshape_alias_226 = None\0A    view_247 = torch.ops.aten.view(sum_129, [384]);  sum_129 = None\0A    detach_315 = torch.ops.aten.detach(view_247);  view_247 = None\0A    detach_316 = torch.ops.aten.detach(detach_315);  detach_315 = None\0A    _reshape_alias_227 = torch.ops.aten._reshape_alias(mm_110, [1, 128, 384], [49152, 384, 1]);  mm_110 = None\0A    add_72 = torch.ops.aten.add(add_71, _reshape_alias_227);  add_71 = _reshape_alias_227 = None\0A    t_297 = torch.ops.aten.t(t_296);  t_296 = None\0A    detach_317 = torch.ops.aten.detach(t_297);  t_297 = None\0A    detach_318 = torch.ops.aten.detach(detach_317);  detach_317 = None\0A    to_162 = torch.ops.aten.to(add_72, 6)\0A    to_163 = torch.ops.aten.to(add_9, 6)\0A    to_164 = torch.ops.aten.to(getitem_19, 6)\0A    to_165 = torch.ops.aten.to(getitem_20, 6)\0A    to_166 = torch.ops.aten.to(params_83, 6)\0A    to_167 = torch.ops.aten.to(params_82, 6)\0A    sub_54 = torch.ops.aten.sub(add_9, getitem_19);  add_9 = getitem_19 = None\0A    mul_126 = torch.ops.aten.mul(sub_54, getitem_20);  sub_54 = None\0A    mul_127 = torch.ops.aten.mul(add_72, params_83)\0A    mul_128 = torch.ops.aten.mul(mul_127, 384)\0A    sum_130 = torch.ops.aten.sum(mul_127, [2], True)\0A    mul_129 = torch.ops.aten.mul(mul_127, mul_126);  mul_127 = None\0A    sum_131 = torch.ops.aten.sum(mul_129, [2], True);  mul_129 = None\0A    mul_130 = torch.ops.aten.mul(mul_126, sum_131);  sum_131 = None\0A    sub_55 = torch.ops.aten.sub(mul_128, sum_130);  mul_128 = sum_130 = None\0A    sub_56 = torch.ops.aten.sub(sub_55, mul_130);  sub_55 = mul_130 = None\0A    div_39 = torch.ops.aten.div(getitem_20, 384);  getitem_20 = None\0A    mul_131 = torch.ops.aten.mul(div_39, sub_56);  div_39 = sub_56 = None\0A    mul_132 = torch.ops.aten.mul(add_72, mul_126);  mul_126 = None\0A    sum_132 = torch.ops.aten.sum(mul_132, [0, 1]);  mul_132 = None\0A    sum_133 = torch.ops.aten.sum(add_72, [0, 1]);  add_72 = None\0A    to_168 = torch.ops.aten.to(mul_131, 6)\0A    to_169 = torch.ops.aten.to(sum_132, 6)\0A    to_170 = torch.ops.aten.to(sum_133, 6)\0A    detach_319 = torch.ops.aten.detach(sum_132);  sum_132 = None\0A    detach_320 = torch.ops.aten.detach(detach_319);  detach_319 = None\0A    detach_321 = torch.ops.aten.detach(sum_133);  sum_133 = None\0A    detach_322 = torch.ops.aten.detach(detach_321);  detach_321 = None\0A    _reshape_alias_228 = torch.ops.aten._reshape_alias(mul_131, [128, 384], [384, 1])\0A    t_298 = torch.ops.aten.t(t_17);  t_17 = None\0A    mm_112 = torch.ops.aten.mm(_reshape_alias_228, t_298);  t_298 = None\0A    t_299 = torch.ops.aten.t(_reshape_alias_228)\0A    mm_113 = torch.ops.aten.mm(t_299, view_46);  t_299 = view_46 = None\0A    t_300 = torch.ops.aten.t(mm_113);  mm_113 = None\0A    sum_134 = torch.ops.aten.sum(_reshape_alias_228, [0], True);  _reshape_alias_228 = None\0A    view_248 = torch.ops.aten.view(sum_134, [384]);  sum_134 = None\0A    detach_323 = torch.ops.aten.detach(view_248);  view_248 = None\0A    detach_324 = torch.ops.aten.detach(detach_323);  detach_323 = None\0A    _reshape_alias_229 = torch.ops.aten._reshape_alias(mm_112, [1, 128, 1536], [196608, 1536, 1]);  mm_112 = None\0A    t_301 = torch.ops.aten.t(t_300);  t_300 = None\0A    detach_325 = torch.ops.aten.detach(t_301);  t_301 = None\0A    detach_326 = torch.ops.aten.detach(detach_325);  detach_325 = None\0A    gelu_backward_9 = torch.ops.aten.gelu_backward(_reshape_alias_229, view_45);  _reshape_alias_229 = view_45 = None\0A    _reshape_alias_230 = torch.ops.aten._reshape_alias(gelu_backward_9, [128, 1536], [1536, 1]);  gelu_backward_9 = None\0A    t_302 = torch.ops.aten.t(t_16);  t_16 = None\0A    mm_114 = torch.ops.aten.mm(_reshape_alias_230, t_302);  t_302 = None\0A    t_303 = torch.ops.aten.t(_reshape_alias_230)\0A    mm_115 = torch.ops.aten.mm(t_303, view_44);  t_303 = view_44 = None\0A    t_304 = torch.ops.aten.t(mm_115);  mm_115 = None\0A    sum_135 = torch.ops.aten.sum(_reshape_alias_230, [0], True);  _reshape_alias_230 = None\0A    view_249 = torch.ops.aten.view(sum_135, [1536]);  sum_135 = None\0A    detach_327 = torch.ops.aten.detach(view_249);  view_249 = None\0A    detach_328 = torch.ops.aten.detach(detach_327);  detach_327 = None\0A    _reshape_alias_231 = torch.ops.aten._reshape_alias(mm_114, [1, 128, 384], [49152, 384, 1]);  mm_114 = None\0A    add_73 = torch.ops.aten.add(mul_131, _reshape_alias_231);  mul_131 = _reshape_alias_231 = None\0A    t_305 = torch.ops.aten.t(t_304);  t_304 = None\0A    detach_329 = torch.ops.aten.detach(t_305);  t_305 = None\0A    detach_330 = torch.ops.aten.detach(detach_329);  detach_329 = None\0A    to_171 = torch.ops.aten.to(add_73, 6)\0A    to_172 = torch.ops.aten.to(add_8, 6)\0A    to_173 = torch.ops.aten.to(getitem_16, 6)\0A    to_174 = torch.ops.aten.to(getitem_17, 6)\0A    to_175 = torch.ops.aten.to(params_71, 6)\0A    to_176 = torch.ops.aten.to(params_70, 6)\0A    sub_57 = torch.ops.aten.sub(add_8, getitem_16);  add_8 = getitem_16 = None\0A    mul_133 = torch.ops.aten.mul(sub_57, getitem_17);  sub_57 = None\0A    mul_134 = torch.ops.aten.mul(add_73, params_71)\0A    mul_135 = torch.ops.aten.mul(mul_134, 384)\0A    sum_136 = torch.ops.aten.sum(mul_134, [2], True)\0A    mul_136 = torch.ops.aten.mul(mul_134, mul_133);  mul_134 = None\0A    sum_137 = torch.ops.aten.sum(mul_136, [2], True);  mul_136 = None\0A    mul_137 = torch.ops.aten.mul(mul_133, sum_137);  sum_137 = None\0A    sub_58 = torch.ops.aten.sub(mul_135, sum_136);  mul_135 = sum_136 = None\0A    sub_59 = torch.ops.aten.sub(sub_58, mul_137);  sub_58 = mul_137 = None\0A    div_40 = torch.ops.aten.div(getitem_17, 384);  getitem_17 = None\0A    mul_138 = torch.ops.aten.mul(div_40, sub_59);  div_40 = sub_59 = None\0A    mul_139 = torch.ops.aten.mul(add_73, mul_133);  mul_133 = None\0A    sum_138 = torch.ops.aten.sum(mul_139, [0, 1]);  mul_139 = None\0A    sum_139 = torch.ops.aten.sum(add_73, [0, 1]);  add_73 = None\0A    to_177 = torch.ops.aten.to(mul_138, 6)\0A    to_178 = torch.ops.aten.to(sum_138, 6)\0A    to_179 = torch.ops.aten.to(sum_139, 6)\0A    detach_331 = torch.ops.aten.detach(sum_138);  sum_138 = None\0A    detach_332 = torch.ops.aten.detach(detach_331);  detach_331 = None\0A    detach_333 = torch.ops.aten.detach(sum_139);  sum_139 = None\0A    detach_334 = torch.ops.aten.detach(detach_333);  detach_333 = None\0A    _reshape_alias_232 = torch.ops.aten._reshape_alias(mul_138, [128, 384], [384, 1])\0A    t_306 = torch.ops.aten.t(t_15);  t_15 = None\0A    mm_116 = torch.ops.aten.mm(_reshape_alias_232, t_306);  t_306 = None\0A    t_307 = torch.ops.aten.t(_reshape_alias_232)\0A    mm_117 = torch.ops.aten.mm(t_307, view_42);  t_307 = view_42 = None\0A    t_308 = torch.ops.aten.t(mm_117);  mm_117 = None\0A    sum_140 = torch.ops.aten.sum(_reshape_alias_232, [0], True);  _reshape_alias_232 = None\0A    view_250 = torch.ops.aten.view(sum_140, [384]);  sum_140 = None\0A    detach_335 = torch.ops.aten.detach(view_250);  view_250 = None\0A    detach_336 = torch.ops.aten.detach(detach_335);  detach_335 = None\0A    _reshape_alias_233 = torch.ops.aten._reshape_alias(mm_116, [1, 128, 384], [49152, 384, 1]);  mm_116 = None\0A    t_309 = torch.ops.aten.t(t_308);  t_308 = None\0A    detach_337 = torch.ops.aten.detach(t_309);  t_309 = None\0A    detach_338 = torch.ops.aten.detach(detach_337);  detach_337 = None\0A    _reshape_alias_234 = torch.ops.aten._reshape_alias(_reshape_alias_233, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_233 = None\0A    permute_84 = torch.ops.aten.permute(_reshape_alias_234, [0, 2, 1, 3]);  _reshape_alias_234 = None\0A    _reshape_alias_235 = torch.ops.aten._reshape_alias(permute_84, [12, 128, 32], [32, 384, 1]);  permute_84 = None\0A    transpose_57 = torch.ops.aten.transpose(_reshape_alias_10, 1, 2);  _reshape_alias_10 = None\0A    bmm_60 = torch.ops.aten.bmm(transpose_57, _reshape_alias_235);  transpose_57 = None\0A    transpose_58 = torch.ops.aten.transpose(_reshape_alias_11, 1, 2);  _reshape_alias_11 = None\0A    bmm_61 = torch.ops.aten.bmm(_reshape_alias_235, transpose_58);  _reshape_alias_235 = transpose_58 = None\0A    _reshape_alias_236 = torch.ops.aten._reshape_alias(bmm_60, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_60 = None\0A    _reshape_alias_237 = torch.ops.aten._reshape_alias(bmm_61, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_61 = None\0A    detach_339 = torch.ops.aten.detach(detach_2);  detach_2 = None\0A    _softmax_backward_data_9 = torch.ops.aten._softmax_backward_data(_reshape_alias_237, detach_339, -1, 6);  _reshape_alias_237 = detach_339 = None\0A    div_41 = torch.ops.aten.div(_softmax_backward_data_9, 5.656854249492381);  _softmax_backward_data_9 = None\0A    _reshape_alias_238 = torch.ops.aten._reshape_alias(div_41, [12, 128, 128], [16384, 128, 1]);  div_41 = None\0A    transpose_59 = torch.ops.aten.transpose(_reshape_alias_8, 1, 2);  _reshape_alias_8 = None\0A    bmm_62 = torch.ops.aten.bmm(transpose_59, _reshape_alias_238);  transpose_59 = None\0A    transpose_60 = torch.ops.aten.transpose(_reshape_alias_9, 1, 2);  _reshape_alias_9 = None\0A    bmm_63 = torch.ops.aten.bmm(_reshape_alias_238, transpose_60);  _reshape_alias_238 = transpose_60 = None\0A    _reshape_alias_239 = torch.ops.aten._reshape_alias(bmm_62, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_62 = None\0A    _reshape_alias_240 = torch.ops.aten._reshape_alias(bmm_63, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_63 = None\0A    transpose_61 = torch.ops.aten.transpose(_reshape_alias_239, -1, -2);  _reshape_alias_239 = None\0A    permute_85 = torch.ops.aten.permute(_reshape_alias_240, [0, 2, 1, 3]);  _reshape_alias_240 = None\0A    clone_30 = torch.ops.aten.clone(permute_85, memory_format = 0);  permute_85 = None\0A    _unsafe_view_42 = torch.ops.aten._unsafe_view(clone_30, [1, 128, 384]);  clone_30 = None\0A    permute_86 = torch.ops.aten.permute(_reshape_alias_236, [0, 2, 1, 3]);  _reshape_alias_236 = None\0A    clone_31 = torch.ops.aten.clone(permute_86, memory_format = 0);  permute_86 = None\0A    _unsafe_view_43 = torch.ops.aten._unsafe_view(clone_31, [1, 128, 384]);  clone_31 = None\0A    _reshape_alias_241 = torch.ops.aten._reshape_alias(_unsafe_view_43, [128, 384], [384, 1]);  _unsafe_view_43 = None\0A    t_310 = torch.ops.aten.t(t_14);  t_14 = None\0A    mm_118 = torch.ops.aten.mm(_reshape_alias_241, t_310);  t_310 = None\0A    t_311 = torch.ops.aten.t(_reshape_alias_241)\0A    mm_119 = torch.ops.aten.mm(t_311, view_37);  t_311 = view_37 = None\0A    t_312 = torch.ops.aten.t(mm_119);  mm_119 = None\0A    sum_141 = torch.ops.aten.sum(_reshape_alias_241, [0], True);  _reshape_alias_241 = None\0A    view_251 = torch.ops.aten.view(sum_141, [384]);  sum_141 = None\0A    detach_340 = torch.ops.aten.detach(view_251);  view_251 = None\0A    detach_341 = torch.ops.aten.detach(detach_340);  detach_340 = None\0A    _reshape_alias_242 = torch.ops.aten._reshape_alias(mm_118, [1, 128, 384], [49152, 384, 1]);  mm_118 = None\0A    add_74 = torch.ops.aten.add(mul_138, _reshape_alias_242);  mul_138 = _reshape_alias_242 = None\0A    t_313 = torch.ops.aten.t(t_312);  t_312 = None\0A    detach_342 = torch.ops.aten.detach(t_313);  t_313 = None\0A    detach_343 = torch.ops.aten.detach(detach_342);  detach_342 = None\0A    permute_87 = torch.ops.aten.permute(transpose_61, [0, 2, 1, 3]);  transpose_61 = None\0A    _reshape_alias_243 = torch.ops.aten._reshape_alias(permute_87, [1, 128, 384], [128, 1, 128]);  permute_87 = None\0A    _reshape_alias_244 = torch.ops.aten._reshape_alias(_reshape_alias_243, [128, 384], [1, 128]);  _reshape_alias_243 = None\0A    t_314 = torch.ops.aten.t(t_13);  t_13 = None\0A    mm_120 = torch.ops.aten.mm(_reshape_alias_244, t_314);  t_314 = None\0A    t_315 = torch.ops.aten.t(_reshape_alias_244)\0A    mm_121 = torch.ops.aten.mm(t_315, view_34);  t_315 = view_34 = None\0A    t_316 = torch.ops.aten.t(mm_121);  mm_121 = None\0A    sum_142 = torch.ops.aten.sum(_reshape_alias_244, [0], True);  _reshape_alias_244 = None\0A    view_252 = torch.ops.aten.view(sum_142, [384]);  sum_142 = None\0A    detach_344 = torch.ops.aten.detach(view_252);  view_252 = None\0A    detach_345 = torch.ops.aten.detach(detach_344);  detach_344 = None\0A    _reshape_alias_245 = torch.ops.aten._reshape_alias(mm_120, [1, 128, 384], [49152, 384, 1]);  mm_120 = None\0A    add_75 = torch.ops.aten.add(add_74, _reshape_alias_245);  add_74 = _reshape_alias_245 = None\0A    t_317 = torch.ops.aten.t(t_316);  t_316 = None\0A    detach_346 = torch.ops.aten.detach(t_317);  t_317 = None\0A    detach_347 = torch.ops.aten.detach(detach_346);  detach_346 = None\0A    _reshape_alias_246 = torch.ops.aten._reshape_alias(_unsafe_view_42, [128, 384], [384, 1]);  _unsafe_view_42 = None\0A    t_318 = torch.ops.aten.t(t_12);  t_12 = None\0A    mm_122 = torch.ops.aten.mm(_reshape_alias_246, t_318);  t_318 = None\0A    t_319 = torch.ops.aten.t(_reshape_alias_246)\0A    mm_123 = torch.ops.aten.mm(t_319, view_32);  t_319 = view_32 = None\0A    t_320 = torch.ops.aten.t(mm_123);  mm_123 = None\0A    sum_143 = torch.ops.aten.sum(_reshape_alias_246, [0], True);  _reshape_alias_246 = None\0A    view_253 = torch.ops.aten.view(sum_143, [384]);  sum_143 = None\0A    detach_348 = torch.ops.aten.detach(view_253);  view_253 = None\0A    detach_349 = torch.ops.aten.detach(detach_348);  detach_348 = None\0A    _reshape_alias_247 = torch.ops.aten._reshape_alias(mm_122, [1, 128, 384], [49152, 384, 1]);  mm_122 = None\0A    add_76 = torch.ops.aten.add(add_75, _reshape_alias_247);  add_75 = _reshape_alias_247 = None\0A    t_321 = torch.ops.aten.t(t_320);  t_320 = None\0A    detach_350 = torch.ops.aten.detach(t_321);  t_321 = None\0A    detach_351 = torch.ops.aten.detach(detach_350);  detach_350 = None\0A    to_180 = torch.ops.aten.to(add_76, 6)\0A    to_181 = torch.ops.aten.to(add_6, 6)\0A    to_182 = torch.ops.aten.to(getitem_13, 6)\0A    to_183 = torch.ops.aten.to(getitem_14, 6)\0A    to_184 = torch.ops.aten.to(params_35, 6)\0A    to_185 = torch.ops.aten.to(params_34, 6)\0A    sub_60 = torch.ops.aten.sub(add_6, getitem_13);  add_6 = getitem_13 = None\0A    mul_140 = torch.ops.aten.mul(sub_60, getitem_14);  sub_60 = None\0A    mul_141 = torch.ops.aten.mul(add_76, params_35)\0A    mul_142 = torch.ops.aten.mul(mul_141, 384)\0A    sum_144 = torch.ops.aten.sum(mul_141, [2], True)\0A    mul_143 = torch.ops.aten.mul(mul_141, mul_140);  mul_141 = None\0A    sum_145 = torch.ops.aten.sum(mul_143, [2], True);  mul_143 = None\0A    mul_144 = torch.ops.aten.mul(mul_140, sum_145);  sum_145 = None\0A    sub_61 = torch.ops.aten.sub(mul_142, sum_144);  mul_142 = sum_144 = None\0A    sub_62 = torch.ops.aten.sub(sub_61, mul_144);  sub_61 = mul_144 = None\0A    div_42 = torch.ops.aten.div(getitem_14, 384);  getitem_14 = None\0A    mul_145 = torch.ops.aten.mul(div_42, sub_62);  div_42 = sub_62 = None\0A    mul_146 = torch.ops.aten.mul(add_76, mul_140);  mul_140 = None\0A    sum_146 = torch.ops.aten.sum(mul_146, [0, 1]);  mul_146 = None\0A    sum_147 = torch.ops.aten.sum(add_76, [0, 1]);  add_76 = None\0A    to_186 = torch.ops.aten.to(mul_145, 6)\0A    to_187 = torch.ops.aten.to(sum_146, 6)\0A    to_188 = torch.ops.aten.to(sum_147, 6)\0A    detach_352 = torch.ops.aten.detach(sum_146);  sum_146 = None\0A    detach_353 = torch.ops.aten.detach(detach_352);  detach_352 = None\0A    detach_354 = torch.ops.aten.detach(sum_147);  sum_147 = None\0A    detach_355 = torch.ops.aten.detach(detach_354);  detach_354 = None\0A    _reshape_alias_248 = torch.ops.aten._reshape_alias(mul_145, [128, 384], [384, 1])\0A    t_322 = torch.ops.aten.t(t_11);  t_11 = None\0A    mm_124 = torch.ops.aten.mm(_reshape_alias_248, t_322);  t_322 = None\0A    t_323 = torch.ops.aten.t(_reshape_alias_248)\0A    mm_125 = torch.ops.aten.mm(t_323, view_30);  t_323 = view_30 = None\0A    t_324 = torch.ops.aten.t(mm_125);  mm_125 = None\0A    sum_148 = torch.ops.aten.sum(_reshape_alias_248, [0], True);  _reshape_alias_248 = None\0A    view_254 = torch.ops.aten.view(sum_148, [384]);  sum_148 = None\0A    detach_356 = torch.ops.aten.detach(view_254);  view_254 = None\0A    detach_357 = torch.ops.aten.detach(detach_356);  detach_356 = None\0A    _reshape_alias_249 = torch.ops.aten._reshape_alias(mm_124, [1, 128, 1536], [196608, 1536, 1]);  mm_124 = None\0A    t_325 = torch.ops.aten.t(t_324);  t_324 = None\0A    detach_358 = torch.ops.aten.detach(t_325);  t_325 = None\0A    detach_359 = torch.ops.aten.detach(detach_358);  detach_358 = None\0A    gelu_backward_10 = torch.ops.aten.gelu_backward(_reshape_alias_249, view_29);  _reshape_alias_249 = view_29 = None\0A    _reshape_alias_250 = torch.ops.aten._reshape_alias(gelu_backward_10, [128, 1536], [1536, 1]);  gelu_backward_10 = None\0A    t_326 = torch.ops.aten.t(t_10);  t_10 = None\0A    mm_126 = torch.ops.aten.mm(_reshape_alias_250, t_326);  t_326 = None\0A    t_327 = torch.ops.aten.t(_reshape_alias_250)\0A    mm_127 = torch.ops.aten.mm(t_327, view_28);  t_327 = view_28 = None\0A    t_328 = torch.ops.aten.t(mm_127);  mm_127 = None\0A    sum_149 = torch.ops.aten.sum(_reshape_alias_250, [0], True);  _reshape_alias_250 = None\0A    view_255 = torch.ops.aten.view(sum_149, [1536]);  sum_149 = None\0A    detach_360 = torch.ops.aten.detach(view_255);  view_255 = None\0A    detach_361 = torch.ops.aten.detach(detach_360);  detach_360 = None\0A    _reshape_alias_251 = torch.ops.aten._reshape_alias(mm_126, [1, 128, 384], [49152, 384, 1]);  mm_126 = None\0A    add_77 = torch.ops.aten.add(mul_145, _reshape_alias_251);  mul_145 = _reshape_alias_251 = None\0A    t_329 = torch.ops.aten.t(t_328);  t_328 = None\0A    detach_362 = torch.ops.aten.detach(t_329);  t_329 = None\0A    detach_363 = torch.ops.aten.detach(detach_362);  detach_362 = None\0A    to_189 = torch.ops.aten.to(add_77, 6)\0A    to_190 = torch.ops.aten.to(add_5, 6)\0A    to_191 = torch.ops.aten.to(getitem_10, 6)\0A    to_192 = torch.ops.aten.to(getitem_11, 6)\0A    to_193 = torch.ops.aten.to(params_23, 6)\0A    to_194 = torch.ops.aten.to(params_22, 6)\0A    sub_63 = torch.ops.aten.sub(add_5, getitem_10);  add_5 = getitem_10 = None\0A    mul_147 = torch.ops.aten.mul(sub_63, getitem_11);  sub_63 = None\0A    mul_148 = torch.ops.aten.mul(add_77, params_23)\0A    mul_149 = torch.ops.aten.mul(mul_148, 384)\0A    sum_150 = torch.ops.aten.sum(mul_148, [2], True)\0A    mul_150 = torch.ops.aten.mul(mul_148, mul_147);  mul_148 = None\0A    sum_151 = torch.ops.aten.sum(mul_150, [2], True);  mul_150 = None\0A    mul_151 = torch.ops.aten.mul(mul_147, sum_151);  sum_151 = None\0A    sub_64 = torch.ops.aten.sub(mul_149, sum_150);  mul_149 = sum_150 = None\0A    sub_65 = torch.ops.aten.sub(sub_64, mul_151);  sub_64 = mul_151 = None\0A    div_43 = torch.ops.aten.div(getitem_11, 384);  getitem_11 = None\0A    mul_152 = torch.ops.aten.mul(div_43, sub_65);  div_43 = sub_65 = None\0A    mul_153 = torch.ops.aten.mul(add_77, mul_147);  mul_147 = None\0A    sum_152 = torch.ops.aten.sum(mul_153, [0, 1]);  mul_153 = None\0A    sum_153 = torch.ops.aten.sum(add_77, [0, 1]);  add_77 = None\0A    to_195 = torch.ops.aten.to(mul_152, 6)\0A    to_196 = torch.ops.aten.to(sum_152, 6)\0A    to_197 = torch.ops.aten.to(sum_153, 6)\0A    detach_364 = torch.ops.aten.detach(sum_152);  sum_152 = None\0A    detach_365 = torch.ops.aten.detach(detach_364);  detach_364 = None\0A    detach_366 = torch.ops.aten.detach(sum_153);  sum_153 = None\0A    detach_367 = torch.ops.aten.detach(detach_366);  detach_366 = None\0A    _reshape_alias_252 = torch.ops.aten._reshape_alias(mul_152, [128, 384], [384, 1])\0A    t_330 = torch.ops.aten.t(t_9);  t_9 = None\0A    mm_128 = torch.ops.aten.mm(_reshape_alias_252, t_330);  t_330 = None\0A    t_331 = torch.ops.aten.t(_reshape_alias_252)\0A    mm_129 = torch.ops.aten.mm(t_331, view_26);  t_331 = view_26 = None\0A    t_332 = torch.ops.aten.t(mm_129);  mm_129 = None\0A    sum_154 = torch.ops.aten.sum(_reshape_alias_252, [0], True);  _reshape_alias_252 = None\0A    view_256 = torch.ops.aten.view(sum_154, [384]);  sum_154 = None\0A    detach_368 = torch.ops.aten.detach(view_256);  view_256 = None\0A    detach_369 = torch.ops.aten.detach(detach_368);  detach_368 = None\0A    _reshape_alias_253 = torch.ops.aten._reshape_alias(mm_128, [1, 128, 384], [49152, 384, 1]);  mm_128 = None\0A    t_333 = torch.ops.aten.t(t_332);  t_332 = None\0A    detach_370 = torch.ops.aten.detach(t_333);  t_333 = None\0A    detach_371 = torch.ops.aten.detach(detach_370);  detach_370 = None\0A    _reshape_alias_254 = torch.ops.aten._reshape_alias(_reshape_alias_253, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_253 = None\0A    permute_88 = torch.ops.aten.permute(_reshape_alias_254, [0, 2, 1, 3]);  _reshape_alias_254 = None\0A    _reshape_alias_255 = torch.ops.aten._reshape_alias(permute_88, [12, 128, 32], [32, 384, 1]);  permute_88 = None\0A    transpose_62 = torch.ops.aten.transpose(_reshape_alias_6, 1, 2);  _reshape_alias_6 = None\0A    bmm_64 = torch.ops.aten.bmm(transpose_62, _reshape_alias_255);  transpose_62 = None\0A    transpose_63 = torch.ops.aten.transpose(_reshape_alias_7, 1, 2);  _reshape_alias_7 = None\0A    bmm_65 = torch.ops.aten.bmm(_reshape_alias_255, transpose_63);  _reshape_alias_255 = transpose_63 = None\0A    _reshape_alias_256 = torch.ops.aten._reshape_alias(bmm_64, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_64 = None\0A    _reshape_alias_257 = torch.ops.aten._reshape_alias(bmm_65, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_65 = None\0A    detach_372 = torch.ops.aten.detach(detach_1);  detach_1 = None\0A    _softmax_backward_data_10 = torch.ops.aten._softmax_backward_data(_reshape_alias_257, detach_372, -1, 6);  _reshape_alias_257 = detach_372 = None\0A    div_44 = torch.ops.aten.div(_softmax_backward_data_10, 5.656854249492381);  _softmax_backward_data_10 = None\0A    _reshape_alias_258 = torch.ops.aten._reshape_alias(div_44, [12, 128, 128], [16384, 128, 1]);  div_44 = None\0A    transpose_64 = torch.ops.aten.transpose(_reshape_alias_4, 1, 2);  _reshape_alias_4 = None\0A    bmm_66 = torch.ops.aten.bmm(transpose_64, _reshape_alias_258);  transpose_64 = None\0A    transpose_65 = torch.ops.aten.transpose(_reshape_alias_5, 1, 2);  _reshape_alias_5 = None\0A    bmm_67 = torch.ops.aten.bmm(_reshape_alias_258, transpose_65);  _reshape_alias_258 = transpose_65 = None\0A    _reshape_alias_259 = torch.ops.aten._reshape_alias(bmm_66, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_66 = None\0A    _reshape_alias_260 = torch.ops.aten._reshape_alias(bmm_67, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_67 = None\0A    transpose_66 = torch.ops.aten.transpose(_reshape_alias_259, -1, -2);  _reshape_alias_259 = None\0A    permute_89 = torch.ops.aten.permute(_reshape_alias_260, [0, 2, 1, 3]);  _reshape_alias_260 = None\0A    clone_32 = torch.ops.aten.clone(permute_89, memory_format = 0);  permute_89 = None\0A    _unsafe_view_44 = torch.ops.aten._unsafe_view(clone_32, [1, 128, 384]);  clone_32 = None\0A    permute_90 = torch.ops.aten.permute(_reshape_alias_256, [0, 2, 1, 3]);  _reshape_alias_256 = None\0A    clone_33 = torch.ops.aten.clone(permute_90, memory_format = 0);  permute_90 = None\0A    _unsafe_view_45 = torch.ops.aten._unsafe_view(clone_33, [1, 128, 384]);  clone_33 = None\0A    _reshape_alias_261 = torch.ops.aten._reshape_alias(_unsafe_view_45, [128, 384], [384, 1]);  _unsafe_view_45 = None\0A    t_334 = torch.ops.aten.t(t_8);  t_8 = None\0A    mm_130 = torch.ops.aten.mm(_reshape_alias_261, t_334);  t_334 = None\0A    t_335 = torch.ops.aten.t(_reshape_alias_261)\0A    mm_131 = torch.ops.aten.mm(t_335, view_21);  t_335 = view_21 = None\0A    t_336 = torch.ops.aten.t(mm_131);  mm_131 = None\0A    sum_155 = torch.ops.aten.sum(_reshape_alias_261, [0], True);  _reshape_alias_261 = None\0A    view_257 = torch.ops.aten.view(sum_155, [384]);  sum_155 = None\0A    detach_373 = torch.ops.aten.detach(view_257);  view_257 = None\0A    detach_374 = torch.ops.aten.detach(detach_373);  detach_373 = None\0A    _reshape_alias_262 = torch.ops.aten._reshape_alias(mm_130, [1, 128, 384], [49152, 384, 1]);  mm_130 = None\0A    add_78 = torch.ops.aten.add(mul_152, _reshape_alias_262);  mul_152 = _reshape_alias_262 = None\0A    t_337 = torch.ops.aten.t(t_336);  t_336 = None\0A    detach_375 = torch.ops.aten.detach(t_337);  t_337 = None\0A    detach_376 = torch.ops.aten.detach(detach_375);  detach_375 = None\0A    permute_91 = torch.ops.aten.permute(transpose_66, [0, 2, 1, 3]);  transpose_66 = None\0A    _reshape_alias_263 = torch.ops.aten._reshape_alias(permute_91, [1, 128, 384], [128, 1, 128]);  permute_91 = None\0A    _reshape_alias_264 = torch.ops.aten._reshape_alias(_reshape_alias_263, [128, 384], [1, 128]);  _reshape_alias_263 = None\0A    t_338 = torch.ops.aten.t(t_7);  t_7 = None\0A    mm_132 = torch.ops.aten.mm(_reshape_alias_264, t_338);  t_338 = None\0A    t_339 = torch.ops.aten.t(_reshape_alias_264)\0A    mm_133 = torch.ops.aten.mm(t_339, view_18);  t_339 = view_18 = None\0A    t_340 = torch.ops.aten.t(mm_133);  mm_133 = None\0A    sum_156 = torch.ops.aten.sum(_reshape_alias_264, [0], True);  _reshape_alias_264 = None\0A    view_258 = torch.ops.aten.view(sum_156, [384]);  sum_156 = None\0A    detach_377 = torch.ops.aten.detach(view_258);  view_258 = None\0A    detach_378 = torch.ops.aten.detach(detach_377);  detach_377 = None\0A    _reshape_alias_265 = torch.ops.aten._reshape_alias(mm_132, [1, 128, 384], [49152, 384, 1]);  mm_132 = None\0A    add_79 = torch.ops.aten.add(add_78, _reshape_alias_265);  add_78 = _reshape_alias_265 = None\0A    t_341 = torch.ops.aten.t(t_340);  t_340 = None\0A    detach_379 = torch.ops.aten.detach(t_341);  t_341 = None\0A    detach_380 = torch.ops.aten.detach(detach_379);  detach_379 = None\0A    _reshape_alias_266 = torch.ops.aten._reshape_alias(_unsafe_view_44, [128, 384], [384, 1]);  _unsafe_view_44 = None\0A    t_342 = torch.ops.aten.t(t_6);  t_6 = None\0A    mm_134 = torch.ops.aten.mm(_reshape_alias_266, t_342);  t_342 = None\0A    t_343 = torch.ops.aten.t(_reshape_alias_266)\0A    mm_135 = torch.ops.aten.mm(t_343, view_16);  t_343 = view_16 = None\0A    t_344 = torch.ops.aten.t(mm_135);  mm_135 = None\0A    sum_157 = torch.ops.aten.sum(_reshape_alias_266, [0], True);  _reshape_alias_266 = None\0A    view_259 = torch.ops.aten.view(sum_157, [384]);  sum_157 = None\0A    detach_381 = torch.ops.aten.detach(view_259);  view_259 = None\0A    detach_382 = torch.ops.aten.detach(detach_381);  detach_381 = None\0A    _reshape_alias_267 = torch.ops.aten._reshape_alias(mm_134, [1, 128, 384], [49152, 384, 1]);  mm_134 = None\0A    add_80 = torch.ops.aten.add(add_79, _reshape_alias_267);  add_79 = _reshape_alias_267 = None\0A    t_345 = torch.ops.aten.t(t_344);  t_344 = None\0A    detach_383 = torch.ops.aten.detach(t_345);  t_345 = None\0A    detach_384 = torch.ops.aten.detach(detach_383);  detach_383 = None\0A    to_198 = torch.ops.aten.to(add_80, 6)\0A    to_199 = torch.ops.aten.to(add_3, 6)\0A    to_200 = torch.ops.aten.to(getitem_7, 6)\0A    to_201 = torch.ops.aten.to(getitem_8, 6)\0A    to_202 = torch.ops.aten.to(params_19, 6)\0A    to_203 = torch.ops.aten.to(params_18, 6)\0A    sub_66 = torch.ops.aten.sub(add_3, getitem_7);  add_3 = getitem_7 = None\0A    mul_154 = torch.ops.aten.mul(sub_66, getitem_8);  sub_66 = None\0A    mul_155 = torch.ops.aten.mul(add_80, params_19)\0A    mul_156 = torch.ops.aten.mul(mul_155, 384)\0A    sum_158 = torch.ops.aten.sum(mul_155, [2], True)\0A    mul_157 = torch.ops.aten.mul(mul_155, mul_154);  mul_155 = None\0A    sum_159 = torch.ops.aten.sum(mul_157, [2], True);  mul_157 = None\0A    mul_158 = torch.ops.aten.mul(mul_154, sum_159);  sum_159 = None\0A    sub_67 = torch.ops.aten.sub(mul_156, sum_158);  mul_156 = sum_158 = None\0A    sub_68 = torch.ops.aten.sub(sub_67, mul_158);  sub_67 = mul_158 = None\0A    div_45 = torch.ops.aten.div(getitem_8, 384);  getitem_8 = None\0A    mul_159 = torch.ops.aten.mul(div_45, sub_68);  div_45 = sub_68 = None\0A    mul_160 = torch.ops.aten.mul(add_80, mul_154);  mul_154 = None\0A    sum_160 = torch.ops.aten.sum(mul_160, [0, 1]);  mul_160 = None\0A    sum_161 = torch.ops.aten.sum(add_80, [0, 1]);  add_80 = None\0A    to_204 = torch.ops.aten.to(mul_159, 6)\0A    to_205 = torch.ops.aten.to(sum_160, 6)\0A    to_206 = torch.ops.aten.to(sum_161, 6)\0A    detach_385 = torch.ops.aten.detach(sum_160);  sum_160 = None\0A    detach_386 = torch.ops.aten.detach(detach_385);  detach_385 = None\0A    detach_387 = torch.ops.aten.detach(sum_161);  sum_161 = None\0A    detach_388 = torch.ops.aten.detach(detach_387);  detach_387 = None\0A    _reshape_alias_268 = torch.ops.aten._reshape_alias(mul_159, [128, 384], [384, 1])\0A    t_346 = torch.ops.aten.t(t_5);  t_5 = None\0A    mm_136 = torch.ops.aten.mm(_reshape_alias_268, t_346);  t_346 = None\0A    t_347 = torch.ops.aten.t(_reshape_alias_268)\0A    mm_137 = torch.ops.aten.mm(t_347, view_14);  t_347 = view_14 = None\0A    t_348 = torch.ops.aten.t(mm_137);  mm_137 = None\0A    sum_162 = torch.ops.aten.sum(_reshape_alias_268, [0], True);  _reshape_alias_268 = None\0A    view_260 = torch.ops.aten.view(sum_162, [384]);  sum_162 = None\0A    detach_389 = torch.ops.aten.detach(view_260);  view_260 = None\0A    detach_390 = torch.ops.aten.detach(detach_389);  detach_389 = None\0A    _reshape_alias_269 = torch.ops.aten._reshape_alias(mm_136, [1, 128, 1536], [196608, 1536, 1]);  mm_136 = None\0A    t_349 = torch.ops.aten.t(t_348);  t_348 = None\0A    detach_391 = torch.ops.aten.detach(t_349);  t_349 = None\0A    detach_392 = torch.ops.aten.detach(detach_391);  detach_391 = None\0A    gelu_backward_11 = torch.ops.aten.gelu_backward(_reshape_alias_269, view_13);  _reshape_alias_269 = view_13 = None\0A    _reshape_alias_270 = torch.ops.aten._reshape_alias(gelu_backward_11, [128, 1536], [1536, 1]);  gelu_backward_11 = None\0A    t_350 = torch.ops.aten.t(t_4);  t_4 = None\0A    mm_138 = torch.ops.aten.mm(_reshape_alias_270, t_350);  t_350 = None\0A    t_351 = torch.ops.aten.t(_reshape_alias_270)\0A    mm_139 = torch.ops.aten.mm(t_351, view_12);  t_351 = view_12 = None\0A    t_352 = torch.ops.aten.t(mm_139);  mm_139 = None\0A    sum_163 = torch.ops.aten.sum(_reshape_alias_270, [0], True);  _reshape_alias_270 = None\0A    view_261 = torch.ops.aten.view(sum_163, [1536]);  sum_163 = None\0A    detach_393 = torch.ops.aten.detach(view_261);  view_261 = None\0A    detach_394 = torch.ops.aten.detach(detach_393);  detach_393 = None\0A    _reshape_alias_271 = torch.ops.aten._reshape_alias(mm_138, [1, 128, 384], [49152, 384, 1]);  mm_138 = None\0A    add_81 = torch.ops.aten.add(mul_159, _reshape_alias_271);  mul_159 = _reshape_alias_271 = None\0A    t_353 = torch.ops.aten.t(t_352);  t_352 = None\0A    detach_395 = torch.ops.aten.detach(t_353);  t_353 = None\0A    detach_396 = torch.ops.aten.detach(detach_395);  detach_395 = None\0A    to_207 = torch.ops.aten.to(add_81, 6)\0A    to_208 = torch.ops.aten.to(add_2, 6)\0A    to_209 = torch.ops.aten.to(getitem_4, 6)\0A    to_210 = torch.ops.aten.to(getitem_5, 6)\0A    to_211 = torch.ops.aten.to(params_7, 6)\0A    to_212 = torch.ops.aten.to(params_6, 6)\0A    sub_69 = torch.ops.aten.sub(add_2, getitem_4);  add_2 = getitem_4 = None\0A    mul_161 = torch.ops.aten.mul(sub_69, getitem_5);  sub_69 = None\0A    mul_162 = torch.ops.aten.mul(add_81, params_7)\0A    mul_163 = torch.ops.aten.mul(mul_162, 384)\0A    sum_164 = torch.ops.aten.sum(mul_162, [2], True)\0A    mul_164 = torch.ops.aten.mul(mul_162, mul_161);  mul_162 = None\0A    sum_165 = torch.ops.aten.sum(mul_164, [2], True);  mul_164 = None\0A    mul_165 = torch.ops.aten.mul(mul_161, sum_165);  sum_165 = None\0A    sub_70 = torch.ops.aten.sub(mul_163, sum_164);  mul_163 = sum_164 = None\0A    sub_71 = torch.ops.aten.sub(sub_70, mul_165);  sub_70 = mul_165 = None\0A    div_46 = torch.ops.aten.div(getitem_5, 384);  getitem_5 = None\0A    mul_166 = torch.ops.aten.mul(div_46, sub_71);  div_46 = sub_71 = None\0A    mul_167 = torch.ops.aten.mul(add_81, mul_161);  mul_161 = None\0A    sum_166 = torch.ops.aten.sum(mul_167, [0, 1]);  mul_167 = None\0A    sum_167 = torch.ops.aten.sum(add_81, [0, 1]);  add_81 = None\0A    to_213 = torch.ops.aten.to(mul_166, 6)\0A    to_214 = torch.ops.aten.to(sum_166, 6)\0A    to_215 = torch.ops.aten.to(sum_167, 6)\0A    detach_397 = torch.ops.aten.detach(sum_166);  sum_166 = None\0A    detach_398 = torch.ops.aten.detach(detach_397);  detach_397 = None\0A    detach_399 = torch.ops.aten.detach(sum_167);  sum_167 = None\0A    detach_400 = torch.ops.aten.detach(detach_399);  detach_399 = None\0A    _reshape_alias_272 = torch.ops.aten._reshape_alias(mul_166, [128, 384], [384, 1])\0A    t_354 = torch.ops.aten.t(t_3);  t_3 = None\0A    mm_140 = torch.ops.aten.mm(_reshape_alias_272, t_354);  t_354 = None\0A    t_355 = torch.ops.aten.t(_reshape_alias_272)\0A    mm_141 = torch.ops.aten.mm(t_355, view_10);  t_355 = view_10 = None\0A    t_356 = torch.ops.aten.t(mm_141);  mm_141 = None\0A    sum_168 = torch.ops.aten.sum(_reshape_alias_272, [0], True);  _reshape_alias_272 = None\0A    view_262 = torch.ops.aten.view(sum_168, [384]);  sum_168 = None\0A    detach_401 = torch.ops.aten.detach(view_262);  view_262 = None\0A    detach_402 = torch.ops.aten.detach(detach_401);  detach_401 = None\0A    _reshape_alias_273 = torch.ops.aten._reshape_alias(mm_140, [1, 128, 384], [49152, 384, 1]);  mm_140 = None\0A    t_357 = torch.ops.aten.t(t_356);  t_356 = None\0A    detach_403 = torch.ops.aten.detach(t_357);  t_357 = None\0A    detach_404 = torch.ops.aten.detach(detach_403);  detach_403 = None\0A    _reshape_alias_274 = torch.ops.aten._reshape_alias(_reshape_alias_273, [1, 128, 12, 32], [49152, 384, 32, 1]);  _reshape_alias_273 = None\0A    permute_92 = torch.ops.aten.permute(_reshape_alias_274, [0, 2, 1, 3]);  _reshape_alias_274 = None\0A    _reshape_alias_275 = torch.ops.aten._reshape_alias(permute_92, [12, 128, 32], [32, 384, 1]);  permute_92 = None\0A    transpose_67 = torch.ops.aten.transpose(_reshape_alias_2, 1, 2);  _reshape_alias_2 = None\0A    bmm_68 = torch.ops.aten.bmm(transpose_67, _reshape_alias_275);  transpose_67 = None\0A    transpose_68 = torch.ops.aten.transpose(_reshape_alias_3, 1, 2);  _reshape_alias_3 = None\0A    bmm_69 = torch.ops.aten.bmm(_reshape_alias_275, transpose_68);  _reshape_alias_275 = transpose_68 = None\0A    _reshape_alias_276 = torch.ops.aten._reshape_alias(bmm_68, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_68 = None\0A    _reshape_alias_277 = torch.ops.aten._reshape_alias(bmm_69, [1, 12, 128, 128], [196608, 16384, 128, 1]);  bmm_69 = None\0A    detach_405 = torch.ops.aten.detach(detach);  detach = None\0A    _softmax_backward_data_11 = torch.ops.aten._softmax_backward_data(_reshape_alias_277, detach_405, -1, 6);  _reshape_alias_277 = detach_405 = None\0A    div_47 = torch.ops.aten.div(_softmax_backward_data_11, 5.656854249492381);  _softmax_backward_data_11 = None\0A    _reshape_alias_278 = torch.ops.aten._reshape_alias(div_47, [12, 128, 128], [16384, 128, 1]);  div_47 = None\0A    transpose_69 = torch.ops.aten.transpose(_reshape_alias, 1, 2);  _reshape_alias = None\0A    bmm_70 = torch.ops.aten.bmm(transpose_69, _reshape_alias_278);  transpose_69 = None\0A    transpose_70 = torch.ops.aten.transpose(_reshape_alias_1, 1, 2);  _reshape_alias_1 = None\0A    bmm_71 = torch.ops.aten.bmm(_reshape_alias_278, transpose_70);  _reshape_alias_278 = transpose_70 = None\0A    _reshape_alias_279 = torch.ops.aten._reshape_alias(bmm_70, [1, 12, 32, 128], [49152, 4096, 128, 1]);  bmm_70 = None\0A    _reshape_alias_280 = torch.ops.aten._reshape_alias(bmm_71, [1, 12, 128, 32], [49152, 4096, 32, 1]);  bmm_71 = None\0A    transpose_71 = torch.ops.aten.transpose(_reshape_alias_279, -1, -2);  _reshape_alias_279 = None\0A    permute_93 = torch.ops.aten.permute(_reshape_alias_280, [0, 2, 1, 3]);  _reshape_alias_280 = None\0A    clone_34 = torch.ops.aten.clone(permute_93, memory_format = 0);  permute_93 = None\0A    _unsafe_view_46 = torch.ops.aten._unsafe_view(clone_34, [1, 128, 384]);  clone_34 = None\0A    permute_94 = torch.ops.aten.permute(_reshape_alias_276, [0, 2, 1, 3]);  _reshape_alias_276 = None\0A    clone_35 = torch.ops.aten.clone(permute_94, memory_format = 0);  permute_94 = None\0A    _unsafe_view_47 = torch.ops.aten._unsafe_view(clone_35, [1, 128, 384]);  clone_35 = None\0A    _reshape_alias_281 = torch.ops.aten._reshape_alias(_unsafe_view_47, [128, 384], [384, 1]);  _unsafe_view_47 = None\0A    t_358 = torch.ops.aten.t(t_2);  t_2 = None\0A    mm_142 = torch.ops.aten.mm(_reshape_alias_281, t_358);  t_358 = None\0A    t_359 = torch.ops.aten.t(_reshape_alias_281)\0A    mm_143 = torch.ops.aten.mm(t_359, view_5);  t_359 = view_5 = None\0A    t_360 = torch.ops.aten.t(mm_143);  mm_143 = None\0A    sum_169 = torch.ops.aten.sum(_reshape_alias_281, [0], True);  _reshape_alias_281 = None\0A    view_263 = torch.ops.aten.view(sum_169, [384]);  sum_169 = None\0A    detach_406 = torch.ops.aten.detach(view_263);  view_263 = None\0A    detach_407 = torch.ops.aten.detach(detach_406);  detach_406 = None\0A    _reshape_alias_282 = torch.ops.aten._reshape_alias(mm_142, [1, 128, 384], [49152, 384, 1]);  mm_142 = None\0A    add_82 = torch.ops.aten.add(mul_166, _reshape_alias_282);  mul_166 = _reshape_alias_282 = None\0A    t_361 = torch.ops.aten.t(t_360);  t_360 = None\0A    detach_408 = torch.ops.aten.detach(t_361);  t_361 = None\0A    detach_409 = torch.ops.aten.detach(detach_408);  detach_408 = None\0A    permute_95 = torch.ops.aten.permute(transpose_71, [0, 2, 1, 3]);  transpose_71 = None\0A    _reshape_alias_283 = torch.ops.aten._reshape_alias(permute_95, [1, 128, 384], [128, 1, 128]);  permute_95 = None\0A    _reshape_alias_284 = torch.ops.aten._reshape_alias(_reshape_alias_283, [128, 384], [1, 128]);  _reshape_alias_283 = None\0A    t_362 = torch.ops.aten.t(t_1);  t_1 = None\0A    mm_144 = torch.ops.aten.mm(_reshape_alias_284, t_362);  t_362 = None\0A    t_363 = torch.ops.aten.t(_reshape_alias_284)\0A    mm_145 = torch.ops.aten.mm(t_363, view_2);  t_363 = view_2 = None\0A    t_364 = torch.ops.aten.t(mm_145);  mm_145 = None\0A    sum_170 = torch.ops.aten.sum(_reshape_alias_284, [0], True);  _reshape_alias_284 = None\0A    view_264 = torch.ops.aten.view(sum_170, [384]);  sum_170 = None\0A    detach_410 = torch.ops.aten.detach(view_264);  view_264 = None\0A    detach_411 = torch.ops.aten.detach(detach_410);  detach_410 = None\0A    _reshape_alias_285 = torch.ops.aten._reshape_alias(mm_144, [1, 128, 384], [49152, 384, 1]);  mm_144 = None\0A    add_83 = torch.ops.aten.add(add_82, _reshape_alias_285);  add_82 = _reshape_alias_285 = None\0A    t_365 = torch.ops.aten.t(t_364);  t_364 = None\0A    detach_412 = torch.ops.aten.detach(t_365);  t_365 = None\0A    detach_413 = torch.ops.aten.detach(detach_412);  detach_412 = None\0A    _reshape_alias_286 = torch.ops.aten._reshape_alias(_unsafe_view_46, [128, 384], [384, 1]);  _unsafe_view_46 = None\0A    t_366 = torch.ops.aten.t(t);  t = None\0A    mm_146 = torch.ops.aten.mm(_reshape_alias_286, t_366);  t_366 = None\0A    t_367 = torch.ops.aten.t(_reshape_alias_286)\0A    mm_147 = torch.ops.aten.mm(t_367, view);  t_367 = view = None\0A    t_368 = torch.ops.aten.t(mm_147);  mm_147 = None\0A    sum_171 = torch.ops.aten.sum(_reshape_alias_286, [0], True);  _reshape_alias_286 = None\0A    view_265 = torch.ops.aten.view(sum_171, [384]);  sum_171 = None\0A    detach_414 = torch.ops.aten.detach(view_265);  view_265 = None\0A    detach_415 = torch.ops.aten.detach(detach_414);  detach_414 = None\0A    _reshape_alias_287 = torch.ops.aten._reshape_alias(mm_146, [1, 128, 384], [49152, 384, 1]);  mm_146 = None\0A    add_84 = torch.ops.aten.add(add_83, _reshape_alias_287);  add_83 = _reshape_alias_287 = None\0A    t_369 = torch.ops.aten.t(t_368);  t_368 = None\0A    detach_416 = torch.ops.aten.detach(t_369);  t_369 = None\0A    detach_417 = torch.ops.aten.detach(detach_416);  detach_416 = None\0A    to_216 = torch.ops.aten.to(add_84, 6)\0A    to_217 = torch.ops.aten.to(add_, 6)\0A    to_218 = torch.ops.aten.to(getitem_1, 6)\0A    to_219 = torch.ops.aten.to(getitem_2, 6)\0A    to_220 = torch.ops.aten.to(params_2, 6)\0A    to_221 = torch.ops.aten.to(params_1, 6)\0A    sub_72 = torch.ops.aten.sub(add_, getitem_1);  add_ = getitem_1 = None\0A    mul_168 = torch.ops.aten.mul(sub_72, getitem_2);  sub_72 = None\0A    mul_169 = torch.ops.aten.mul(add_84, params_2)\0A    mul_170 = torch.ops.aten.mul(mul_169, 384)\0A    sum_172 = torch.ops.aten.sum(mul_169, [2], True)\0A    mul_171 = torch.ops.aten.mul(mul_169, mul_168);  mul_169 = None\0A    sum_173 = torch.ops.aten.sum(mul_171, [2], True);  mul_171 = None\0A    mul_172 = torch.ops.aten.mul(mul_168, sum_173);  sum_173 = None\0A    sub_73 = torch.ops.aten.sub(mul_170, sum_172);  mul_170 = sum_172 = None\0A    sub_74 = torch.ops.aten.sub(sub_73, mul_172);  sub_73 = mul_172 = None\0A    div_48 = torch.ops.aten.div(getitem_2, 384);  getitem_2 = None\0A    mul_173 = torch.ops.aten.mul(div_48, sub_74);  div_48 = sub_74 = None\0A    mul_174 = torch.ops.aten.mul(add_84, mul_168);  mul_168 = None\0A    sum_174 = torch.ops.aten.sum(mul_174, [0, 1]);  mul_174 = None\0A    sum_175 = torch.ops.aten.sum(add_84, [0, 1]);  add_84 = None\0A    to_222 = torch.ops.aten.to(mul_173, 6)\0A    to_223 = torch.ops.aten.to(sum_174, 6)\0A    to_224 = torch.ops.aten.to(sum_175, 6)\0A    detach_418 = torch.ops.aten.detach(sum_174);  sum_174 = None\0A    detach_419 = torch.ops.aten.detach(detach_418);  detach_418 = None\0A    detach_420 = torch.ops.aten.detach(sum_175);  sum_175 = None\0A    detach_421 = torch.ops.aten.detach(detach_420);  detach_420 = None\0A    view_266 = torch.ops.aten.view(mul_173, [128, 384])\0A    new_zeros_2 = torch.ops.aten.new_zeros(mul_173, [512, 384], device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    view_267 = torch.ops.aten.view(slice_4, [128]);  slice_4 = None\0A    ne = torch.ops.aten.ne(view_267, -1)\0A    unsqueeze = torch.ops.aten.unsqueeze(ne, 1);  ne = None\0A    expand_as = torch.ops.aten.expand_as(unsqueeze, view_266);  unsqueeze = None\0A    full_like = torch.ops.aten.full_like(view_266, 0, device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    where = torch.ops.aten.where(expand_as, view_266, full_like);  expand_as = view_266 = full_like = None\0A    index_put = torch.ops.aten.index_put(new_zeros_2, [view_267], where, True);  new_zeros_2 = view_267 = where = None\0A    detach_422 = torch.ops.aten.detach(index_put);  index_put = None\0A    detach_423 = torch.ops.aten.detach(detach_422);  detach_422 = None\0A    view_268 = torch.ops.aten.view(mul_173, [128, 384])\0A    new_zeros_3 = torch.ops.aten.new_zeros(mul_173, [2, 384], device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    view_269 = torch.ops.aten.view(expand, [128]);  expand = None\0A    ne_1 = torch.ops.aten.ne(view_269, -1)\0A    unsqueeze_1 = torch.ops.aten.unsqueeze(ne_1, 1);  ne_1 = None\0A    expand_as_1 = torch.ops.aten.expand_as(unsqueeze_1, view_268);  unsqueeze_1 = None\0A    full_like_1 = torch.ops.aten.full_like(view_268, 0, device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    where_1 = torch.ops.aten.where(expand_as_1, view_268, full_like_1);  expand_as_1 = view_268 = full_like_1 = None\0A    index_put_1 = torch.ops.aten.index_put(new_zeros_3, [view_269], where_1, True);  new_zeros_3 = view_269 = where_1 = None\0A    detach_424 = torch.ops.aten.detach(index_put_1);  index_put_1 = None\0A    detach_425 = torch.ops.aten.detach(detach_424);  detach_424 = None\0A    view_270 = torch.ops.aten.view(mul_173, [128, 384])\0A    new_zeros_4 = torch.ops.aten.new_zeros(mul_173, [30522, 384], device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False);  mul_173 = None\0A    view_271 = torch.ops.aten.view(args_1, [128]);  args_1 = None\0A    ne_2 = torch.ops.aten.ne(view_271, 0)\0A    unsqueeze_2 = torch.ops.aten.unsqueeze(ne_2, 1);  ne_2 = None\0A    expand_as_2 = torch.ops.aten.expand_as(unsqueeze_2, view_270);  unsqueeze_2 = None\0A    full_like_2 = torch.ops.aten.full_like(view_270, 0, device = device(type='cpu'), dtype = 6, layout = 0, pin_memory = False)\0A    where_2 = torch.ops.aten.where(expand_as_2, view_270, full_like_2);  expand_as_2 = view_270 = full_like_2 = None\0A    index_put_2 = torch.ops.aten.index_put(new_zeros_4, [view_271], where_2, True);  new_zeros_4 = view_271 = where_2 = None\0A    detach_426 = torch.ops.aten.detach(index_put_2);  index_put_2 = None\0A    detach_427 = torch.ops.aten.detach(detach_426);  detach_426 = None\0A    add__1 = torch.ops.aten.add_(params_1, detach_421, alpha = -0.01);  params_1 = detach_421 = None\0A    add__2 = torch.ops.aten.add_(params_2, detach_419, alpha = -0.01);  params_2 = detach_419 = None\0A    add__3 = torch.ops.aten.add_(params_3, detach_423, alpha = -0.01);  params_3 = detach_423 = None\0A    add__4 = torch.ops.aten.add_(params_4, detach_425, alpha = -0.01);  params_4 = detach_425 = None\0A    add__5 = torch.ops.aten.add_(params_5, detach_427, alpha = -0.01);  params_5 = detach_427 = None\0A    add__6 = torch.ops.aten.add_(params_6, detach_400, alpha = -0.01);  params_6 = detach_400 = None\0A    add__7 = torch.ops.aten.add_(params_7, detach_398, alpha = -0.01);  params_7 = detach_398 = None\0A    add__8 = torch.ops.aten.add_(params_8, detach_402, alpha = -0.01);  params_8 = detach_402 = None\0A    add__9 = torch.ops.aten.add_(params_9, detach_404, alpha = -0.01);  params_9 = detach_404 = None\0A    add__10 = torch.ops.aten.add_(params_10, detach_411, alpha = -0.01);  params_10 = detach_411 = None\0A    add__11 = torch.ops.aten.add_(params_11, detach_413, alpha = -0.01);  params_11 = detach_413 = None\0A    add__12 = torch.ops.aten.add_(params_12, detach_415, alpha = -0.01);  params_12 = detach_415 = None\0A    add__13 = torch.ops.aten.add_(params_13, detach_417, alpha = -0.01);  params_13 = detach_417 = None\0A    add__14 = torch.ops.aten.add_(params_14, detach_407, alpha = -0.01);  params_14 = detach_407 = None\0A    add__15 = torch.ops.aten.add_(params_15, detach_409, alpha = -0.01);  params_15 = detach_409 = None\0A    add__16 = torch.ops.aten.add_(params_16, detach_394, alpha = -0.01);  params_16 = detach_394 = None\0A    add__17 = torch.ops.aten.add_(params_17, detach_396, alpha = -0.01);  params_17 = detach_396 = None\0A    add__18 = torch.ops.aten.add_(params_18, detach_388, alpha = -0.01);  params_18 = detach_388 = None\0A    add__19 = torch.ops.aten.add_(params_19, detach_386, alpha = -0.01);  params_19 = detach_386 = None\0A    add__20 = torch.ops.aten.add_(params_20, detach_390, alpha = -0.01);  params_20 = detach_390 = None\0A    add__21 = torch.ops.aten.add_(params_21, detach_392, alpha = -0.01);  params_21 = detach_392 = None\0A    add__22 = torch.ops.aten.add_(params_22, detach_367, alpha = -0.01);  params_22 = detach_367 = None\0A    add__23 = torch.ops.aten.add_(params_23, detach_365, alpha = -0.01);  params_23 = detach_365 = None\0A    add__24 = torch.ops.aten.add_(params_24, detach_369, alpha = -0.01);  params_24 = detach_369 = None\0A    add__25 = torch.ops.aten.add_(params_25, detach_371, alpha = -0.01);  params_25 = detach_371 = None\0A    add__26 = torch.ops.aten.add_(params_26, detach_378, alpha = -0.01);  params_26 = detach_378 = None\0A    add__27 = torch.ops.aten.add_(params_27, detach_380, alpha = -0.01);  params_27 = detach_380 = None\0A    add__28 = torch.ops.aten.add_(params_28, detach_382, alpha = -0.01);  params_28 = detach_382 = None\0A    add__29 = torch.ops.aten.add_(params_29, detach_384, alpha = -0.01);  params_29 = detach_384 = None\0A    add__30 = torch.ops.aten.add_(params_30, detach_374, alpha = -0.01);  params_30 = detach_374 = None\0A    add__31 = torch.ops.aten.add_(params_31, detach_376, alpha = -0.01);  params_31 = detach_376 = None\0A    add__32 = torch.ops.aten.add_(params_32, detach_361, alpha = -0.01);  params_32 = detach_361 = None\0A    add__33 = torch.ops.aten.add_(params_33, detach_363, alpha = -0.01);  params_33 = detach_363 = None\0A    add__34 = torch.ops.aten.add_(params_34, detach_355, alpha = -0.01);  params_34 = detach_355 = None\0A    add__35 = torch.ops.aten.add_(params_35, detach_353, alpha = -0.01);  params_35 = detach_353 = None\0A    add__36 = torch.ops.aten.add_(params_36, detach_357, alpha = -0.01);  params_36 = detach_357 = None\0A    add__37 = torch.ops.aten.add_(params_37, detach_359, alpha = -0.01);  params_37 = detach_359 = None\0A    add__38 = torch.ops.aten.add_(params_38, detach_70, alpha = -0.01);  params_38 = detach_70 = None\0A    add__39 = torch.ops.aten.add_(params_39, detach_68, alpha = -0.01);  params_39 = detach_68 = None\0A    add__40 = torch.ops.aten.add_(params_40, detach_72, alpha = -0.01);  params_40 = detach_72 = None\0A    add__41 = torch.ops.aten.add_(params_41, detach_74, alpha = -0.01);  params_41 = detach_74 = None\0A    add__42 = torch.ops.aten.add_(params_42, detach_81, alpha = -0.01);  params_42 = detach_81 = None\0A    add__43 = torch.ops.aten.add_(params_43, detach_83, alpha = -0.01);  params_43 = detach_83 = None\0A    add__44 = torch.ops.aten.add_(params_44, detach_85, alpha = -0.01);  params_44 = detach_85 = None\0A    add__45 = torch.ops.aten.add_(params_45, detach_87, alpha = -0.01);  params_45 = detach_87 = None\0A    add__46 = torch.ops.aten.add_(params_46, detach_77, alpha = -0.01);  params_46 = detach_77 = None\0A    add__47 = torch.ops.aten.add_(params_47, detach_79, alpha = -0.01);  params_47 = detach_79 = None\0A    add__48 = torch.ops.aten.add_(params_48, detach_64, alpha = -0.01);  params_48 = detach_64 = None\0A    add__49 = torch.ops.aten.add_(params_49, detach_66, alpha = -0.01);  params_49 = detach_66 = None\0A    add__50 = torch.ops.aten.add_(params_50, detach_58, alpha = -0.01);  params_50 = detach_58 = None\0A    add__51 = torch.ops.aten.add_(params_51, detach_56, alpha = -0.01);  params_51 = detach_56 = None\0A    add__52 = torch.ops.aten.add_(params_52, detach_60, alpha = -0.01);  params_52 = detach_60 = None\0A    add__53 = torch.ops.aten.add_(params_53, detach_62, alpha = -0.01);  params_53 = detach_62 = None\0A    add__54 = torch.ops.aten.add_(params_54, detach_37, alpha = -0.01);  params_54 = detach_37 = None\0A    add__55 = torch.ops.aten.add_(params_55, detach_35, alpha = -0.01);  params_55 = detach_35 = None\0A    add__56 = torch.ops.aten.add_(params_56, detach_39, alpha = -0.01);  params_56 = detach_39 = None\0A    add__57 = torch.ops.aten.add_(params_57, detach_41, alpha = -0.01);  params_57 = detach_41 = None\0A    add__58 = torch.ops.aten.add_(params_58, detach_48, alpha = -0.01);  params_58 = detach_48 = None\0A    add__59 = torch.ops.aten.add_(params_59, detach_50, alpha = -0.01);  params_59 = detach_50 = None\0A    add__60 = torch.ops.aten.add_(params_60, detach_52, alpha = -0.01);  params_60 = detach_52 = None\0A    add__61 = torch.ops.aten.add_(params_61, detach_54, alpha = -0.01);  params_61 = detach_54 = None\0A    add__62 = torch.ops.aten.add_(params_62, detach_44, alpha = -0.01);  params_62 = detach_44 = None\0A    add__63 = torch.ops.aten.add_(params_63, detach_46, alpha = -0.01);  params_63 = detach_46 = None\0A    add__64 = torch.ops.aten.add_(params_64, detach_31, alpha = -0.01);  params_64 = detach_31 = None\0A    add__65 = torch.ops.aten.add_(params_65, detach_33, alpha = -0.01);  params_65 = detach_33 = None\0A    add__66 = torch.ops.aten.add_(params_66, detach_25, alpha = -0.01);  params_66 = detach_25 = None\0A    add__67 = torch.ops.aten.add_(params_67, detach_23, alpha = -0.01);  params_67 = detach_23 = None\0A    add__68 = torch.ops.aten.add_(params_68, detach_27, alpha = -0.01);  params_68 = detach_27 = None\0A    add__69 = torch.ops.aten.add_(params_69, detach_29, alpha = -0.01);  params_69 = detach_29 = None\0A    add__70 = torch.ops.aten.add_(params_70, detach_334, alpha = -0.01);  params_70 = detach_334 = None\0A    add__71 = torch.ops.aten.add_(params_71, detach_332, alpha = -0.01);  params_71 = detach_332 = None\0A    add__72 = torch.ops.aten.add_(params_72, detach_336, alpha = -0.01);  params_72 = detach_336 = None\0A    add__73 = torch.ops.aten.add_(params_73, detach_338, alpha = -0.01);  params_73 = detach_338 = None\0A    add__74 = torch.ops.aten.add_(params_74, detach_345, alpha = -0.01);  params_74 = detach_345 = None\0A    add__75 = torch.ops.aten.add_(params_75, detach_347, alpha = -0.01);  params_75 = detach_347 = None\0A    add__76 = torch.ops.aten.add_(params_76, detach_349, alpha = -0.01);  params_76 = detach_349 = None\0A    add__77 = torch.ops.aten.add_(params_77, detach_351, alpha = -0.01);  params_77 = detach_351 = None\0A    add__78 = torch.ops.aten.add_(params_78, detach_341, alpha = -0.01);  params_78 = detach_341 = None\0A    add__79 = torch.ops.aten.add_(params_79, detach_343, alpha = -0.01);  params_79 = detach_343 = None\0A    add__80 = torch.ops.aten.add_(params_80, detach_328, alpha = -0.01);  params_80 = detach_328 = None\0A    add__81 = torch.ops.aten.add_(params_81, detach_330, alpha = -0.01);  params_81 = detach_330 = None\0A    add__82 = torch.ops.aten.add_(params_82, detach_322, alpha = -0.01);  params_82 = detach_322 = None\0A    add__83 = torch.ops.aten.add_(params_83, detach_320, alpha = -0.01);  params_83 = detach_320 = None\0A    add__84 = torch.ops.aten.add_(params_84, detach_324, alpha = -0.01);  params_84 = detach_324 = None\0A    add__85 = torch.ops.aten.add_(params_85, detach_326, alpha = -0.01);  params_85 = detach_326 = None\0A    add__86 = torch.ops.aten.add_(params_86, detach_301, alpha = -0.01);  params_86 = detach_301 = None\0A    add__87 = torch.ops.aten.add_(params_87, detach_299, alpha = -0.01);  params_87 = detach_299 = None\0A    add__88 = torch.ops.aten.add_(params_88, detach_303, alpha = -0.01);  params_88 = detach_303 = None\0A    add__89 = torch.ops.aten.add_(params_89, detach_305, alpha = -0.01);  params_89 = detach_305 = None\0A    add__90 = torch.ops.aten.add_(params_90, detach_312, alpha = -0.01);  params_90 = detach_312 = None\0A    add__91 = torch.ops.aten.add_(params_91, detach_314, alpha = -0.01);  params_91 = detach_314 = None\0A    add__92 = torch.ops.aten.add_(params_92, detach_316, alpha = -0.01);  params_92 = detach_316 = None\0A    add__93 = torch.ops.aten.add_(params_93, detach_318, alpha = -0.01);  params_93 = detach_318 = None\0A    add__94 = torch.ops.aten.add_(params_94, detach_308, alpha = -0.01);  params_94 = detach_308 = None\0A    add__95 = torch.ops.aten.add_(params_95, detach_310, alpha = -0.01);  params_95 = detach_310 = None\0A    add__96 = torch.ops.aten.add_(params_96, detach_295, alpha = -0.01);  params_96 = detach_295 = None\0A    add__97 = torch.ops.aten.add_(params_97, detach_297, alpha = -0.01);  params_97 = detach_297 = None\0A    add__98 = torch.ops.aten.add_(params_98, detach_289, alpha = -0.01);  params_98 = detach_289 = None\0A    add__99 = torch.ops.aten.add_(params_99, detach_287, alpha = -0.01);  params_99 = detach_287 = None\0A    add__100 = torch.ops.aten.add_(params_100, detach_291, alpha = -0.01);  params_100 = detach_291 = None\0A    add__101 = torch.ops.aten.add_(params_101, detach_293, alpha = -0.01);  params_101 = detach_293 = None\0A    add__102 = torch.ops.aten.add_(params_102, detach_268, alpha = -0.01);  params_102 = detach_268 = None\0A    add__103 = torch.ops.aten.add_(params_103, detach_266, alpha = -0.01);  params_103 = detach_266 = None\0A    add__104 = torch.ops.aten.add_(params_104, detach_270, alpha = -0.01);  params_104 = detach_270 = None\0A    add__105 = torch.ops.aten.add_(params_105, detach_272, alpha = -0.01);  params_105 = detach_272 = None\0A    add__106 = torch.ops.aten.add_(params_106, detach_279, alpha = -0.01);  params_106 = detach_279 = None\0A    add__107 = torch.ops.aten.add_(params_107, detach_281, alpha = -0.01);  params_107 = detach_281 = None\0A    add__108 = torch.ops.aten.add_(params_108, detach_283, alpha = -0.01);  params_108 = detach_283 = None\0A    add__109 = torch.ops.aten.add_(params_109, detach_285, alpha = -0.01);  params_109 = detach_285 = None\0A    add__110 = torch.ops.aten.add_(params_110, detach_275, alpha = -0.01);  params_110 = detach_275 = None\0A    add__111 = torch.ops.aten.add_(params_111, detach_277, alpha = -0.01);  params_111 = detach_277 = None\0A    add__112 = torch.ops.aten.add_(params_112, detach_262, alpha = -0.01);  params_112 = detach_262 = None\0A    add__113 = torch.ops.aten.add_(params_113, detach_264, alpha = -0.01);  params_113 = detach_264 = None\0A    add__114 = torch.ops.aten.add_(params_114, detach_256, alpha = -0.01);  params_114 = detach_256 = None\0A    add__115 = torch.ops.aten.add_(params_115, detach_254, alpha = -0.01);  params_115 = detach_254 = None\0A    add__116 = torch.ops.aten.add_(params_116, detach_258, alpha = -0.01);  params_116 = detach_258 = None\0A    add__117 = torch.ops.aten.add_(params_117, detach_260, alpha = -0.01);  params_117 = detach_260 = None\0A    add__118 = torch.ops.aten.add_(params_118, detach_235, alpha = -0.01);  params_118 = detach_235 = None\0A    add__119 = torch.ops.aten.add_(params_119, detach_233, alpha = -0.01);  params_119 = detach_233 = None\0A    add__120 = torch.ops.aten.add_(params_120, detach_237, alpha = -0.01);  params_120 = detach_237 = None\0A    add__121 = torch.ops.aten.add_(params_121, detach_239, alpha = -0.01);  params_121 = detach_239 = None\0A    add__122 = torch.ops.aten.add_(params_122, detach_246, alpha = -0.01);  params_122 = detach_246 = None\0A    add__123 = torch.ops.aten.add_(params_123, detach_248, alpha = -0.01);  params_123 = detach_248 = None\0A    add__124 = torch.ops.aten.add_(params_124, detach_250, alpha = -0.01);  params_124 = detach_250 = None\0A    add__125 = torch.ops.aten.add_(params_125, detach_252, alpha = -0.01);  params_125 = detach_252 = None\0A    add__126 = torch.ops.aten.add_(params_126, detach_242, alpha = -0.01);  params_126 = detach_242 = None\0A    add__127 = torch.ops.aten.add_(params_127, detach_244, alpha = -0.01);  params_127 = detach_244 = None\0A    add__128 = torch.ops.aten.add_(params_128, detach_229, alpha = -0.01);  params_128 = detach_229 = None\0A    add__129 = torch.ops.aten.add_(params_129, detach_231, alpha = -0.01);  params_129 = detach_231 = None\0A    add__130 = torch.ops.aten.add_(params_130, detach_223, alpha = -0.01);  params_130 = detach_223 = None\0A    add__131 = torch.ops.aten.add_(params_131, detach_221, alpha = -0.01);  params_131 = detach_221 = None\0A    add__132 = torch.ops.aten.add_(params_132, detach_225, alpha = -0.01);  params_132 = detach_225 = None\0A    add__133 = torch.ops.aten.add_(params_133, detach_227, alpha = -0.01);  params_133 = detach_227 = None\0A    add__134 = torch.ops.aten.add_(params_134, detach_202, alpha = -0.01);  params_134 = detach_202 = None\0A    add__135 = torch.ops.aten.add_(params_135, detach_200, alpha = -0.01);  params_135 = detach_200 = None\0A    add__136 = torch.ops.aten.add_(params_136, detach_204, alpha = -0.01);  params_136 = detach_204 = None\0A    add__137 = torch.ops.aten.add_(params_137, detach_206, alpha = -0.01);  params_137 = detach_206 = None\0A    add__138 = torch.ops.aten.add_(params_138, detach_213, alpha = -0.01);  params_138 = detach_213 = None\0A    add__139 = torch.ops.aten.add_(params_139, detach_215, alpha = -0.01);  params_139 = detach_215 = None\0A    add__140 = torch.ops.aten.add_(params_140, detach_217, alpha = -0.01);  params_140 = detach_217 = None\0A    add__141 = torch.ops.aten.add_(params_141, detach_219, alpha = -0.01);  params_141 = detach_219 = None\0A    add__142 = torch.ops.aten.add_(params_142, detach_209, alpha = -0.01);  params_142 = detach_209 = None\0A    add__143 = torch.ops.aten.add_(params_143, detach_211, alpha = -0.01);  params_143 = detach_211 = None\0A    add__144 = torch.ops.aten.add_(params_144, detach_196, alpha = -0.01);  params_144 = detach_196 = None\0A    add__145 = torch.ops.aten.add_(params_145, detach_198, alpha = -0.01);  params_145 = detach_198 = None\0A    add__146 = torch.ops.aten.add_(params_146, detach_190, alpha = -0.01);  params_146 = detach_190 = None\0A    add__147 = torch.ops.aten.add_(params_147, detach_188, alpha = -0.01);  params_147 = detach_188 = None\0A    add__148 = torch.ops.aten.add_(params_148, detach_192, alpha = -0.01);  params_148 = detach_192 = None\0A    add__149 = torch.ops.aten.add_(params_149, detach_194, alpha = -0.01);  params_149 = detach_194 = None\0A    add__150 = torch.ops.aten.add_(params_150, detach_169, alpha = -0.01);  params_150 = detach_169 = None\0A    add__151 = torch.ops.aten.add_(params_151, detach_167, alpha = -0.01);  params_151 = detach_167 = None\0A    add__152 = torch.ops.aten.add_(params_152, detach_171, alpha = -0.01);  params_152 = detach_171 = None\0A    add__153 = torch.ops.aten.add_(params_153, detach_173, alpha = -0.01);  params_153 = detach_173 = None\0A    add__154 = torch.ops.aten.add_(params_154, detach_180, alpha = -0.01);  params_154 = detach_180 = None\0A    add__155 = torch.ops.aten.add_(params_155, detach_182, alpha = -0.01);  params_155 = detach_182 = None\0A    add__156 = torch.ops.aten.add_(params_156, detach_184, alpha = -0.01);  params_156 = detach_184 = None\0A    add__157 = torch.ops.aten.add_(params_157, detach_186, alpha = -0.01);  params_157 = detach_186 = None\0A    add__158 = torch.ops.aten.add_(params_158, detach_176, alpha = -0.01);  params_158 = detach_176 = None\0A    add__159 = torch.ops.aten.add_(params_159, detach_178, alpha = -0.01);  params_159 = detach_178 = None\0A    add__160 = torch.ops.aten.add_(params_160, detach_163, alpha = -0.01);  params_160 = detach_163 = None\0A    add__161 = torch.ops.aten.add_(params_161, detach_165, alpha = -0.01);  params_161 = detach_165 = None\0A    add__162 = torch.ops.aten.add_(params_162, detach_157, alpha = -0.01);  params_162 = detach_157 = None\0A    add__163 = torch.ops.aten.add_(params_163, detach_155, alpha = -0.01);  params_163 = detach_155 = None\0A    add__164 = torch.ops.aten.add_(params_164, detach_159, alpha = -0.01);  params_164 = detach_159 = None\0A    add__165 = torch.ops.aten.add_(params_165, detach_161, alpha = -0.01);  params_165 = detach_161 = None\0A    add__166 = torch.ops.aten.add_(params_166, detach_136, alpha = -0.01);  params_166 = detach_136 = None\0A    add__167 = torch.ops.aten.add_(params_167, detach_134, alpha = -0.01);  params_167 = detach_134 = None\0A    add__168 = torch.ops.aten.add_(params_168, detach_138, alpha = -0.01);  params_168 = detach_138 = None\0A    add__169 = torch.ops.aten.add_(params_169, detach_140, alpha = -0.01);  params_169 = detach_140 = None\0A    add__170 = torch.ops.aten.add_(params_170, detach_147, alpha = -0.01);  params_170 = detach_147 = None\0A    add__171 = torch.ops.aten.add_(params_171, detach_149, alpha = -0.01);  params_171 = detach_149 = None\0A    add__172 = torch.ops.aten.add_(params_172, detach_151, alpha = -0.01);  params_172 = detach_151 = None\0A    add__173 = torch.ops.aten.add_(params_173, detach_153, alpha = -0.01);  params_173 = detach_153 = None\0A    add__174 = torch.ops.aten.add_(params_174, detach_143, alpha = -0.01);  params_174 = detach_143 = None\0A    add__175 = torch.ops.aten.add_(params_175, detach_145, alpha = -0.01);  params_175 = detach_145 = None\0A    add__176 = torch.ops.aten.add_(params_176, detach_130, alpha = -0.01);  params_176 = detach_130 = None\0A    add__177 = torch.ops.aten.add_(params_177, detach_132, alpha = -0.01);  params_177 = detach_132 = None\0A    add__178 = torch.ops.aten.add_(params_178, detach_124, alpha = -0.01);  params_178 = detach_124 = None\0A    add__179 = torch.ops.aten.add_(params_179, detach_122, alpha = -0.01);  params_179 = detach_122 = None\0A    add__180 = torch.ops.aten.add_(params_180, detach_126, alpha = -0.01);  params_180 = detach_126 = None\0A    add__181 = torch.ops.aten.add_(params_181, detach_128, alpha = -0.01);  params_181 = detach_128 = None\0A    add__182 = torch.ops.aten.add_(params_182, detach_103, alpha = -0.01);  params_182 = detach_103 = None\0A    add__183 = torch.ops.aten.add_(params_183, detach_101, alpha = -0.01);  params_183 = detach_101 = None\0A    add__184 = torch.ops.aten.add_(params_184, detach_105, alpha = -0.01);  params_184 = detach_105 = None\0A    add__185 = torch.ops.aten.add_(params_185, detach_107, alpha = -0.01);  params_185 = detach_107 = None\0A    add__186 = torch.ops.aten.add_(params_186, detach_114, alpha = -0.01);  params_186 = detach_114 = None\0A    add__187 = torch.ops.aten.add_(params_187, detach_116, alpha = -0.01);  params_187 = detach_116 = None\0A    add__188 = torch.ops.aten.add_(params_188, detach_118, alpha = -0.01);  params_188 = detach_118 = None\0A    add__189 = torch.ops.aten.add_(params_189, detach_120, alpha = -0.01);  params_189 = detach_120 = None\0A    add__190 = torch.ops.aten.add_(params_190, detach_110, alpha = -0.01);  params_190 = detach_110 = None\0A    add__191 = torch.ops.aten.add_(params_191, detach_112, alpha = -0.01);  params_191 = detach_112 = None\0A    add__192 = torch.ops.aten.add_(params_192, detach_97, alpha = -0.01);  params_192 = detach_97 = None\0A    add__193 = torch.ops.aten.add_(params_193, detach_99, alpha = -0.01);  params_193 = detach_99 = None\0A    add__194 = torch.ops.aten.add_(params_194, detach_91, alpha = -0.01);  params_194 = detach_91 = None\0A    add__195 = torch.ops.aten.add_(params_195, detach_89, alpha = -0.01);  params_195 = detach_89 = None\0A    add__196 = torch.ops.aten.add_(params_196, detach_93, alpha = -0.01);  params_196 = detach_93 = None\0A    add__197 = torch.ops.aten.add_(params_197, detach_95, alpha = -0.01);  params_197 = detach_95 = None\0A    add__198 = torch.ops.aten.add_(params_198, detach_19, alpha = -0.01);  params_198 = detach_19 = None\0A    add__199 = torch.ops.aten.add_(params_199, detach_21, alpha = -0.01);  params_199 = detach_21 = None\0A    add__200 = torch.ops.aten.add_(params_200, detach_14, alpha = -0.01);  params_200 = detach_14 = None\0A    add__201 = torch.ops.aten.add_(params_201, detach_16, alpha = -0.01);  params_201 = detach_16 = None\0A    return (add__1, add__2, add__3, add__4, add__5, add__6, add__7, add__8, add__9, add__10, add__11, add__12, add__13, add__14, add__15, add__16, add__17, add__18, add__19, add__20, add__21, add__22, add__23, add__24, add__25, add__26, add__27, add__28, add__29, add__30, add__31, add__32, add__33, add__34, add__35, add__36, add__37, add__38, add__39, add__40, add__41, add__42, add__43, add__44, add__45, add__46, add__47, add__48, add__49, add__50, add__51, add__52, add__53, add__54, add__55, add__56, add__57, add__58, add__59, add__60, add__61, add__62, add__63, add__64, add__65, add__66, add__67, add__68, add__69, add__70, add__71, add__72, add__73, add__74, add__75, add__76, add__77, add__78, add__79, add__80, add__81, add__82, add__83, add__84, add__85, add__86, add__87, add__88, add__89, add__90, add__91, add__92, add__93, add__94, add__95, add__96, add__97, add__98, add__99, add__100, add__101, add__102, add__103, add__104, add__105, add__106, add__107, add__108, add__109, add__110, add__111, add__112, add__113, add__114, add__115, add__116, add__117, add__118, add__119, add__120, add__121, add__122, add__123, add__124, add__125, add__126, add__127, add__128, add__129, add__130, add__131, add__132, add__133, add__134, add__135, add__136, add__137, add__138, add__139, add__140, add__141, add__142, add__143, add__144, add__145, add__146, add__147, add__148, add__149, add__150, add__151, add__152, add__153, add__154, add__155, add__156, add__157, add__158, add__159, add__160, add__161, add__162, add__163, add__164, add__165, add__166, add__167, add__168, add__169, add__170, add__171, add__172, add__173, add__174, add__175, add__176, add__177, add__178, add__179, add__180, add__181, add__182, add__183, add__184, add__185, add__186, add__187, add__188, add__189, add__190, add__191, add__192, add__193, add__194, add__195, add__196, add__197, add__198, add__199, add__200, add__201, buffers_1, buffers_2)\0A    "
  %1 = torch.nn_module {
    torch.slot "_tensor_constant0", %0 : !torch.tensor<[1,1,1,128],f32>
    torch.slot "training", %true : !torch.bool
    torch.slot "_is_full_backward_hook", %none : !torch.none
    torch.slot "_code", %str : !torch.str
  } : !torch.nn.Module<"__torch__.torch.fx.graph_module.forward">
}