pashu123 · April 5, 2022 15:40
diff --git a/backward-decomposed.mlir b/backward-decomposed.mlir
 module attributes {torch.debug_module_name = "GraphModule"} {
  func private @__torch__.torch.fx.graph_module.___torch_mangle_2.GraphModule.forward(%arg0: !torch.nn.Module<"__torch__.torch.fx.graph_module.___torch_mangle_2.GraphModule">, %arg1: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg2: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg3: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg4: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg5: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg6: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg7: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg8: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg9: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg10: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg11: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg12: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg13: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg14: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg15: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg16: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg17: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg18: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg19: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg20: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg21: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg22: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg23: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg24: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg25: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg26: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg27: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg28: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg29: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg30: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg31: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg32: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg33: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg34: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg35: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg36: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg37: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg38: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg39: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg40: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg41: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg42: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg43: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg44: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg45: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg46: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg47: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg48: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg49: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg50: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg51: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg52: !torch.tensor {torch.type_bound = !torch.vtensor<[768],f32>}, %arg53: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512],si64>}, %arg54: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512],si64>}, %arg55: !torch.tensor {torch.type_bound = !torch.vtensor<[1,512],si64>}, %arg56: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg57: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg58: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg59: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg60: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg61: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg62: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg63: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg64: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg65: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg66: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg67: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg68: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg69: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg70: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg71: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg72: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg73: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg74: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg75: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg76: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg77: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg78: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg79: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg80: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg81: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg82: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg83: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg84: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg85: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg86: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg87: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg88: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg89: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg90: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg91: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg92: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg93: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg94: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg95: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg96: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg97: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg98: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg99: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg100: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg101: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg102: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg103: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg104: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg105: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg106: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg107: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg108: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg109: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg110: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg111: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg112: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg113: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg114: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg115: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg116: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg117: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg118: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg119: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg120: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg121: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg122: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg123: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg124: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg125: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg126: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg127: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg128: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg129: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg130: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg131: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg132: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg133: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg134: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg135: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg136: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg137: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg138: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg139: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg140: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg141: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg142: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg143: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg144: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg145: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg146: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg147: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg148: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg149: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg150: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg151: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg152: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg153: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg154: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg155: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg156: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg157: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg158: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg159: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg160: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg161: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg162: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg163: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg164: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg165: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg166: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg167: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg168: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg169: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg170: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg171: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg172: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg173: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg174: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg175: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg176: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg177: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg178: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg179: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg180: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg181: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg182: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg183: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg184: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg185: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg186: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg187: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg188: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg189: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg190: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg191: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg192: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg193: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg194: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg195: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg196: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg197: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg198: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg199: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg200: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg201: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg202: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg203: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg204: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg205: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg206: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg207: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg208: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg209: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg210: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg211: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg212: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg213: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg214: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg215: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg216: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg217: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg218: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg219: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg220: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg221: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg222: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg223: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg224: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg225: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg226: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg227: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg228: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg229: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg230: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg231: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg232: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg233: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg234: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg235: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg236: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg237: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg238: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg239: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg240: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg241: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg242: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg243: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg244: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg245: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg246: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg247: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg248: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg249: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg250: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg251: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg252: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg253: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg254: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg255: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg256: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg257: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg258: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg259: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg260: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg261: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg262: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg263: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg264: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg265: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg266: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg267: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg268: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg269: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg270: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg271: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg272: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg273: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg274: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg275: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg276: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg277: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg278: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg279: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg280: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg281: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg282: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg283: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg284: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg285: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg286: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg287: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg288: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg289: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg290: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg291: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg292: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg293: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg294: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg295: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg296: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg297: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg298: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg299: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg300: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg301: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg302: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg303: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg304: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg305: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg306: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg307: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg308: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg309: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg310: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg311: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg312: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg313: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg314: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg315: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg316: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg317: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg318: !torch.tensor {torch.type_bound = !torch.vtensor<[48,64,512],f32>}, %arg319: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg320: !torch.tensor {torch.type_bound = !torch.vtensor<[4,12,512,512],f32>}, %arg321: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,512],f32>}, %arg322: !torch.tensor {torch.type_bound = !torch.vtensor<[48,512,64],f32>}, %arg323: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg324: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg325: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg326: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg327: !torch.tensor {torch.type_bound = !torch.vtensor<[768,3072],f32>}, %arg328: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg329: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,3072],f32>}, %arg330: !torch.tensor {torch.type_bound = !torch.vtensor<[3072,768],f32>}, %arg331: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,3072],f32>}, %arg332: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg333: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg334: !torch.tensor {torch.type_bound = !torch.vtensor<[768,768],f32>}, %arg335: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg336: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg337: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,768],f32>}, %arg338: !torch.tensor {torch.type_bound = !torch.vtensor<[768,30522],f32>}, %arg339: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,768],f32>}, %arg340: !torch.tensor {torch.type_bound = !torch.vtensor<[2048],si64>}, %arg341: !torch.tensor {torch.type_bound = !torch.vtensor<[2048,30522],f32>}, %arg342: !torch.tensor {torch.type_bound = !torch.vtensor<[],f32>}, %arg343: !torch.tensor {torch.type_bound = !torch.vtensor<[4,512,30522],f32>}) -> !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, tensor, tensor, tensor, tensor, tensor> {
    %cpu = torch.constant.device "cpu"
    %1 = torch.tensor.literal(dense<8.000000e+00> : tensor<f64>) : !torch.tensor<[],f64>
    %str = torch.constant.str "none"
    %false = torch.constant.bool false
    %2 = torch.tensor.literal(dense<768> : tensor<si64>) : !torch.tensor<[],si64>
    %true = torch.constant.bool true
    %none = torch.constant.none
    %int768 = torch.constant.int 768
    %float9.999990e-13 = torch.constant.float 9.9999999999999998E-13
    %int1 = torch.constant.int 1
    %int-100 = torch.constant.int -100
    %int6 = torch.constant.int 6
    %int4 = torch.constant.int 4
    %int512 = torch.constant.int 512
    %int30522 = torch.constant.int 30522
    %int2048 = torch.constant.int 2048
    %int0 = torch.constant.int 0
    %int2 = torch.constant.int 2
    %int3072 = torch.constant.int 3072
    %int12 = torch.constant.int 12
    %int64 = torch.constant.int 64
    %int3 = torch.constant.int 3
    %int48 = torch.constant.int 48
    %int-1 = torch.constant.int -1
    %int-2 = torch.constant.int -2
    %3 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0, %result1, %result2 = torch.aten.native_layer_norm %arg56, %3, %arg2, %arg1, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %4 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_0, %result1_1, %result2_2 = torch.aten.native_layer_norm %arg73, %4, %arg4, %arg3, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %5 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_3, %result1_4, %result2_5 = torch.aten.native_layer_norm %arg80, %5, %arg6, %arg5, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %6 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_6, %result1_7, %result2_8 = torch.aten.native_layer_norm %arg96, %6, %arg8, %arg7, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %7 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_9, %result1_10, %result2_11 = torch.aten.native_layer_norm %arg103, %7, %arg10, %arg9, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %8 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_12, %result1_13, %result2_14 = torch.aten.native_layer_norm %arg119, %8, %arg20, %arg19, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %9 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_15, %result1_16, %result2_17 = torch.aten.native_layer_norm %arg126, %9, %arg22, %arg21, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %10 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_18, %result1_19, %result2_20 = torch.aten.native_layer_norm %arg142, %10, %arg24, %arg23, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %11 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_21, %result1_22, %result2_23 = torch.aten.native_layer_norm %arg149, %11, %arg26, %arg25, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %12 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_24, %result1_25, %result2_26 = torch.aten.native_layer_norm %arg165, %12, %arg28, %arg27, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %13 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_27, %result1_28, %result2_29 = torch.aten.native_layer_norm %arg172, %13, %arg30, %arg29, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %14 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_30, %result1_31, %result2_32 = torch.aten.native_layer_norm %arg188, %14, %arg32, %arg31, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %15 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_33, %result1_34, %result2_35 = torch.aten.native_layer_norm %arg195, %15, %arg34, %arg33, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %16 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_36, %result1_37, %result2_38 = torch.aten.native_layer_norm %arg211, %16, %arg36, %arg35, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %17 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_39, %result1_40, %result2_41 = torch.aten.native_layer_norm %arg218, %17, %arg38, %arg37, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %18 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_42, %result1_43, %result2_44 = torch.aten.native_layer_norm %arg234, %18, %arg40, %arg39, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %19 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_45, %result1_46, %result2_47 = torch.aten.native_layer_norm %arg241, %19, %arg42, %arg41, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %20 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_48, %result1_49, %result2_50 = torch.aten.native_layer_norm %arg257, %20, %arg44, %arg43, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %21 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_51, %result1_52, %result2_53 = torch.aten.native_layer_norm %arg264, %21, %arg46, %arg45, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %22 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_54, %result1_55, %result2_56 = torch.aten.native_layer_norm %arg280, %22, %arg48, %arg47, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %23 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_57, %result1_58, %result2_59 = torch.aten.native_layer_norm %arg287, %23, %arg50, %arg49, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %24 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_60, %result1_61, %result2_62 = torch.aten.native_layer_norm %arg303, %24, %arg12, %arg11, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %25 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_63, %result1_64, %result2_65 = torch.aten.native_layer_norm %arg310, %25, %arg14, %arg13, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %26 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_66, %result1_67, %result2_68 = torch.aten.native_layer_norm %arg326, %26, %arg16, %arg15, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %27 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_69, %result1_70, %result2_71 = torch.aten.native_layer_norm %arg333, %27, %arg18, %arg17, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %28 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %result0_72, %result1_73, %result2_74 = torch.aten.native_layer_norm %arg337, %28, %arg52, %arg51, %float9.999990e-13 : !torch.tensor, !torch.list<int>, !torch.tensor, !torch.tensor, !torch.float -> !torch.tensor, !torch.tensor, !torch.tensor
    %output, %total_weight = torch.aten.nll_loss_forward %arg341, %arg340, %none, %int1, %int-100 : !torch.tensor, !torch.tensor, !torch.none, !torch.int, !torch.int -> !torch.tensor, !torch.tensor
    %29 = torch.aten.nll_loss_backward %arg342, %arg341, %arg340, %none, %int1, %int-100, %total_weight : !torch.tensor, !torch.tensor, !torch.tensor, !torch.none, !torch.int, !torch.int, !torch.tensor -> !torch.tensor
    %30 = torch.aten._log_softmax_backward_data %29, %arg341, %int1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %31 = torch.prim.ListConstruct %int4, %int512, %int30522 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %32 = torch.aten.view %30, %31 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %non = torch.aten.view %30, %31 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %33 = torch.aten.add.Tensor %arg343, %32, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %34 = torch.prim.ListConstruct %int2048, %int30522 : (!torch.int, !torch.int) -> !torch.list<int>
    %35 = torch.aten.view %33, %34 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %36 = torch.aten.t %arg338 : !torch.tensor -> !torch.tensor
    %37 = torch.aten.mm %35, %36 : !torch.tensor, !torch.tensor -> !torch.tensor
    %38 = torch.aten.t %35 : !torch.tensor -> !torch.tensor
    %39 = torch.aten.mm %38, %arg339 : !torch.tensor, !torch.tensor -> !torch.tensor
    %40 = torch.aten.t %39 : !torch.tensor -> !torch.tensor
    %41 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %42 = torch.aten.sum.dim_IntList %35, %41, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %43 = torch.prim.ListConstruct %int30522 : (!torch.int) -> !torch.list<int>
    %44 = torch.aten.view %42, %43 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %45 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %46 = torch.aten.view %37, %45 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %47 = torch.aten.t %40 : !torch.tensor -> !torch.tensor
    %48 = torch.aten.sub.Tensor %arg337, %result1_73, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %49 = torch.aten.mul.Tensor %48, %result2_74 : !torch.tensor, !torch.tensor -> !torch.tensor
    %50 = torch.aten.mul.Tensor %46, %arg52 : !torch.tensor, !torch.tensor -> !torch.tensor
    %51 = torch.aten.mul.Tensor %50, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %52 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %53 = torch.aten.sum.dim_IntList %50, %52, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %54 = torch.aten.mul.Tensor %50, %49 : !torch.tensor, !torch.tensor -> !torch.tensor
    %55 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %56 = torch.aten.sum.dim_IntList %54, %55, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %57 = torch.aten.mul.Tensor %49, %56 : !torch.tensor, !torch.tensor -> !torch.tensor
    %58 = torch.aten.sub.Tensor %51, %53, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %59 = torch.aten.sub.Tensor %58, %57, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %60 = torch.aten.div.Tensor %result2_74, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %61 = torch.aten.mul.Tensor %60, %59 : !torch.tensor, !torch.tensor -> !torch.tensor
    %62 = torch.aten.mul.Tensor %46, %49 : !torch.tensor, !torch.tensor -> !torch.tensor
    %63 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %64 = torch.aten.sum.dim_IntList %62, %63, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %65 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %66 = torch.aten.sum.dim_IntList %46, %65, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %67 = torch.aten.gelu_backward %61, %arg336, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %68 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %69 = torch.aten.view %67, %68 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %70 = torch.aten.t %arg334 : !torch.tensor -> !torch.tensor
    %71 = torch.aten.mm %69, %70 : !torch.tensor, !torch.tensor -> !torch.tensor
    %72 = torch.aten.t %69 : !torch.tensor -> !torch.tensor
    %73 = torch.aten.mm %72, %arg335 : !torch.tensor, !torch.tensor -> !torch.tensor
    %74 = torch.aten.t %73 : !torch.tensor -> !torch.tensor
    %75 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %76 = torch.aten.sum.dim_IntList %69, %75, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %77 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %78 = torch.aten.view %76, %77 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %79 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %80 = torch.aten.view %71, %79 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %81 = torch.aten.t %74 : !torch.tensor -> !torch.tensor
    %82 = torch.aten.sub.Tensor %arg333, %result1_70, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %83 = torch.aten.mul.Tensor %82, %result2_71 : !torch.tensor, !torch.tensor -> !torch.tensor
    %84 = torch.aten.mul.Tensor %80, %arg18 : !torch.tensor, !torch.tensor -> !torch.tensor
    %85 = torch.aten.mul.Tensor %84, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %86 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %87 = torch.aten.sum.dim_IntList %84, %86, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %88 = torch.aten.mul.Tensor %84, %83 : !torch.tensor, !torch.tensor -> !torch.tensor
    %89 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %90 = torch.aten.sum.dim_IntList %88, %89, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %91 = torch.aten.mul.Tensor %83, %90 : !torch.tensor, !torch.tensor -> !torch.tensor
    %92 = torch.aten.sub.Tensor %85, %87, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %93 = torch.aten.sub.Tensor %92, %91, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %94 = torch.aten.div.Tensor %result2_71, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %95 = torch.aten.mul.Tensor %94, %93 : !torch.tensor, !torch.tensor -> !torch.tensor
    %96 = torch.aten.mul.Tensor %80, %83 : !torch.tensor, !torch.tensor -> !torch.tensor
    %97 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %98 = torch.aten.sum.dim_IntList %96, %97, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %99 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %100 = torch.aten.sum.dim_IntList %80, %99, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %101 = torch.aten.mul.Tensor %95, %arg332 : !torch.tensor, !torch.tensor -> !torch.tensor
    %102 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %103 = torch.aten.view %101, %102 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %104 = torch.aten.t %arg330 : !torch.tensor -> !torch.tensor
    %105 = torch.aten.mm %103, %104 : !torch.tensor, !torch.tensor -> !torch.tensor
    %106 = torch.aten.t %103 : !torch.tensor -> !torch.tensor
    %107 = torch.aten.mm %106, %arg331 : !torch.tensor, !torch.tensor -> !torch.tensor
    %108 = torch.aten.t %107 : !torch.tensor -> !torch.tensor
    %109 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %110 = torch.aten.sum.dim_IntList %103, %109, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %111 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %112 = torch.aten.view %110, %111 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %113 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %114 = torch.aten.view %105, %113 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %115 = torch.aten.t %108 : !torch.tensor -> !torch.tensor
    %116 = torch.aten.gelu_backward %114, %arg329, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %117 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %118 = torch.aten.view %116, %117 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %119 = torch.aten.t %arg327 : !torch.tensor -> !torch.tensor
    %120 = torch.aten.mm %118, %119 : !torch.tensor, !torch.tensor -> !torch.tensor
    %121 = torch.aten.t %118 : !torch.tensor -> !torch.tensor
    %122 = torch.aten.mm %121, %arg328 : !torch.tensor, !torch.tensor -> !torch.tensor
    %123 = torch.aten.t %122 : !torch.tensor -> !torch.tensor
    %124 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %125 = torch.aten.sum.dim_IntList %118, %124, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %126 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %127 = torch.aten.view %125, %126 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %128 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %129 = torch.aten.view %120, %128 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %130 = torch.aten.add.Tensor %95, %129, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %131 = torch.aten.t %123 : !torch.tensor -> !torch.tensor
    %132 = torch.aten.sub.Tensor %arg326, %result1_67, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %133 = torch.aten.mul.Tensor %132, %result2_68 : !torch.tensor, !torch.tensor -> !torch.tensor
    %134 = torch.aten.mul.Tensor %130, %arg16 : !torch.tensor, !torch.tensor -> !torch.tensor
    %135 = torch.aten.mul.Tensor %134, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %136 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %137 = torch.aten.sum.dim_IntList %134, %136, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %138 = torch.aten.mul.Tensor %134, %133 : !torch.tensor, !torch.tensor -> !torch.tensor
    %139 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %140 = torch.aten.sum.dim_IntList %138, %139, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %141 = torch.aten.mul.Tensor %133, %140 : !torch.tensor, !torch.tensor -> !torch.tensor
    %142 = torch.aten.sub.Tensor %135, %137, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %143 = torch.aten.sub.Tensor %142, %141, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %144 = torch.aten.div.Tensor %result2_68, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %145 = torch.aten.mul.Tensor %144, %143 : !torch.tensor, !torch.tensor -> !torch.tensor
    %146 = torch.aten.mul.Tensor %130, %133 : !torch.tensor, !torch.tensor -> !torch.tensor
    %147 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %148 = torch.aten.sum.dim_IntList %146, %147, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %149 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %150 = torch.aten.sum.dim_IntList %130, %149, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %151 = torch.aten.mul.Tensor %145, %arg325 : !torch.tensor, !torch.tensor -> !torch.tensor
    %152 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %153 = torch.aten.view %151, %152 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %154 = torch.aten.t %arg323 : !torch.tensor -> !torch.tensor
    %155 = torch.aten.mm %153, %154 : !torch.tensor, !torch.tensor -> !torch.tensor
    %156 = torch.aten.t %153 : !torch.tensor -> !torch.tensor
    %157 = torch.aten.mm %156, %arg324 : !torch.tensor, !torch.tensor -> !torch.tensor
    %158 = torch.aten.t %157 : !torch.tensor -> !torch.tensor
    %159 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %160 = torch.aten.sum.dim_IntList %153, %159, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %161 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %162 = torch.aten.view %160, %161 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %163 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %164 = torch.aten.view %155, %163 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %165 = torch.aten.t %158 : !torch.tensor -> !torch.tensor
    %166 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %167 = torch.aten.view %164, %166 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %168 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %169 = torch.aten.permute %167, %168 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %170 = torch.aten.clone %169, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %171 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %172 = torch.aten._unsafe_view %170, %171 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %173 = torch.aten.transpose.int %arg321, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %174 = torch.aten.bmm %173, %172 : !torch.tensor, !torch.tensor -> !torch.tensor
    %175 = torch.aten.transpose.int %arg322, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %176 = torch.aten.bmm %172, %175 : !torch.tensor, !torch.tensor -> !torch.tensor
    %177 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %178 = torch.aten.view %174, %177 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %179 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %180 = torch.aten.view %176, %179 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %181 = torch.aten.mul.Tensor %180, %arg320 : !torch.tensor, !torch.tensor -> !torch.tensor
    %182 = torch.aten._softmax_backward_data %181, %arg319, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %183 = torch.aten.div.Tensor %182, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %184 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %185 = torch.aten.view %183, %184 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %186 = torch.aten.transpose.int %arg317, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %187 = torch.aten.bmm %186, %185 : !torch.tensor, !torch.tensor -> !torch.tensor
    %188 = torch.aten.transpose.int %arg318, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %189 = torch.aten.bmm %185, %188 : !torch.tensor, !torch.tensor -> !torch.tensor
    %190 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %191 = torch.aten.view %187, %190 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %192 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %193 = torch.aten.view %189, %192 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %194 = torch.aten.transpose.int %191, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %195 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %196 = torch.aten.permute %193, %195 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %197 = torch.aten.clone %196, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %198 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %199 = torch.aten._unsafe_view %197, %198 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %200 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %201 = torch.aten.permute %178, %200 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %202 = torch.aten.clone %201, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %203 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %204 = torch.aten._unsafe_view %202, %203 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %205 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %206 = torch.aten.view %204, %205 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %207 = torch.aten.t %arg315 : !torch.tensor -> !torch.tensor
    %208 = torch.aten.mm %206, %207 : !torch.tensor, !torch.tensor -> !torch.tensor
    %209 = torch.aten.t %206 : !torch.tensor -> !torch.tensor
    %210 = torch.aten.mm %209, %arg316 : !torch.tensor, !torch.tensor -> !torch.tensor
    %211 = torch.aten.t %210 : !torch.tensor -> !torch.tensor
    %212 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %213 = torch.aten.sum.dim_IntList %206, %212, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %214 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %215 = torch.aten.view %213, %214 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %216 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %217 = torch.aten.view %208, %216 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %218 = torch.aten.add.Tensor %145, %217, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %219 = torch.aten.t %211 : !torch.tensor -> !torch.tensor
    %220 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %221 = torch.aten.permute %194, %220 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %222 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %223 = torch.aten.view %221, %222 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %224 = torch.aten.clone %223, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %225 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %226 = torch.aten._unsafe_view %224, %225 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %227 = torch.aten.t %arg313 : !torch.tensor -> !torch.tensor
    %228 = torch.aten.mm %226, %227 : !torch.tensor, !torch.tensor -> !torch.tensor
    %229 = torch.aten.t %226 : !torch.tensor -> !torch.tensor
    %230 = torch.aten.mm %229, %arg314 : !torch.tensor, !torch.tensor -> !torch.tensor
    %231 = torch.aten.t %230 : !torch.tensor -> !torch.tensor
    %232 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %233 = torch.aten.sum.dim_IntList %226, %232, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %234 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %235 = torch.aten.view %233, %234 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %236 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %237 = torch.aten.view %228, %236 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %238 = torch.aten.add.Tensor %218, %237, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %239 = torch.aten.t %231 : !torch.tensor -> !torch.tensor
    %240 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %241 = torch.aten.view %199, %240 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %242 = torch.aten.t %arg311 : !torch.tensor -> !torch.tensor
    %243 = torch.aten.mm %241, %242 : !torch.tensor, !torch.tensor -> !torch.tensor
    %244 = torch.aten.t %241 : !torch.tensor -> !torch.tensor
    %245 = torch.aten.mm %244, %arg312 : !torch.tensor, !torch.tensor -> !torch.tensor
    %246 = torch.aten.t %245 : !torch.tensor -> !torch.tensor
    %247 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %248 = torch.aten.sum.dim_IntList %241, %247, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %249 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %250 = torch.aten.view %248, %249 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %251 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %252 = torch.aten.view %243, %251 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %253 = torch.aten.add.Tensor %238, %252, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %254 = torch.aten.t %246 : !torch.tensor -> !torch.tensor
    %255 = torch.aten.sub.Tensor %arg310, %result1_64, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %256 = torch.aten.mul.Tensor %255, %result2_65 : !torch.tensor, !torch.tensor -> !torch.tensor
    %257 = torch.aten.mul.Tensor %253, %arg14 : !torch.tensor, !torch.tensor -> !torch.tensor
    %258 = torch.aten.mul.Tensor %257, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %259 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %260 = torch.aten.sum.dim_IntList %257, %259, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %261 = torch.aten.mul.Tensor %257, %256 : !torch.tensor, !torch.tensor -> !torch.tensor
    %262 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %263 = torch.aten.sum.dim_IntList %261, %262, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %264 = torch.aten.mul.Tensor %256, %263 : !torch.tensor, !torch.tensor -> !torch.tensor
    %265 = torch.aten.sub.Tensor %258, %260, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %266 = torch.aten.sub.Tensor %265, %264, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %267 = torch.aten.div.Tensor %result2_65, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %268 = torch.aten.mul.Tensor %267, %266 : !torch.tensor, !torch.tensor -> !torch.tensor
    %269 = torch.aten.mul.Tensor %253, %256 : !torch.tensor, !torch.tensor -> !torch.tensor
    %270 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %271 = torch.aten.sum.dim_IntList %269, %270, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %272 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %273 = torch.aten.sum.dim_IntList %253, %272, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %274 = torch.aten.mul.Tensor %268, %arg309 : !torch.tensor, !torch.tensor -> !torch.tensor
    %275 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %276 = torch.aten.view %274, %275 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %277 = torch.aten.t %arg307 : !torch.tensor -> !torch.tensor
    %278 = torch.aten.mm %276, %277 : !torch.tensor, !torch.tensor -> !torch.tensor
    %279 = torch.aten.t %276 : !torch.tensor -> !torch.tensor
    %280 = torch.aten.mm %279, %arg308 : !torch.tensor, !torch.tensor -> !torch.tensor
    %281 = torch.aten.t %280 : !torch.tensor -> !torch.tensor
    %282 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %283 = torch.aten.sum.dim_IntList %276, %282, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %284 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %285 = torch.aten.view %283, %284 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %286 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %287 = torch.aten.view %278, %286 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %288 = torch.aten.t %281 : !torch.tensor -> !torch.tensor
    %289 = torch.aten.gelu_backward %287, %arg306, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %290 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %291 = torch.aten.view %289, %290 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %292 = torch.aten.t %arg304 : !torch.tensor -> !torch.tensor
    %293 = torch.aten.mm %291, %292 : !torch.tensor, !torch.tensor -> !torch.tensor
    %294 = torch.aten.t %291 : !torch.tensor -> !torch.tensor
    %295 = torch.aten.mm %294, %arg305 : !torch.tensor, !torch.tensor -> !torch.tensor
    %296 = torch.aten.t %295 : !torch.tensor -> !torch.tensor
    %297 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %298 = torch.aten.sum.dim_IntList %291, %297, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %299 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %300 = torch.aten.view %298, %299 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %301 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %302 = torch.aten.view %293, %301 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %303 = torch.aten.add.Tensor %268, %302, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %304 = torch.aten.t %296 : !torch.tensor -> !torch.tensor
    %305 = torch.aten.sub.Tensor %arg303, %result1_61, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %306 = torch.aten.mul.Tensor %305, %result2_62 : !torch.tensor, !torch.tensor -> !torch.tensor
    %307 = torch.aten.mul.Tensor %303, %arg12 : !torch.tensor, !torch.tensor -> !torch.tensor
    %308 = torch.aten.mul.Tensor %307, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %309 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %310 = torch.aten.sum.dim_IntList %307, %309, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %311 = torch.aten.mul.Tensor %307, %306 : !torch.tensor, !torch.tensor -> !torch.tensor
    %312 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %313 = torch.aten.sum.dim_IntList %311, %312, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %314 = torch.aten.mul.Tensor %306, %313 : !torch.tensor, !torch.tensor -> !torch.tensor
    %315 = torch.aten.sub.Tensor %308, %310, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %316 = torch.aten.sub.Tensor %315, %314, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %317 = torch.aten.div.Tensor %result2_62, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %318 = torch.aten.mul.Tensor %317, %316 : !torch.tensor, !torch.tensor -> !torch.tensor
    %319 = torch.aten.mul.Tensor %303, %306 : !torch.tensor, !torch.tensor -> !torch.tensor
    %320 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %321 = torch.aten.sum.dim_IntList %319, %320, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %322 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %323 = torch.aten.sum.dim_IntList %303, %322, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %324 = torch.aten.mul.Tensor %318, %arg302 : !torch.tensor, !torch.tensor -> !torch.tensor
    %325 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %326 = torch.aten.view %324, %325 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %327 = torch.aten.t %arg300 : !torch.tensor -> !torch.tensor
    %328 = torch.aten.mm %326, %327 : !torch.tensor, !torch.tensor -> !torch.tensor
    %329 = torch.aten.t %326 : !torch.tensor -> !torch.tensor
    %330 = torch.aten.mm %329, %arg301 : !torch.tensor, !torch.tensor -> !torch.tensor
    %331 = torch.aten.t %330 : !torch.tensor -> !torch.tensor
    %332 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %333 = torch.aten.sum.dim_IntList %326, %332, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %334 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %335 = torch.aten.view %333, %334 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %336 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %337 = torch.aten.view %328, %336 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %338 = torch.aten.t %331 : !torch.tensor -> !torch.tensor
    %339 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %340 = torch.aten.view %337, %339 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %341 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %342 = torch.aten.permute %340, %341 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %343 = torch.aten.clone %342, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %344 = torch.prim.ListConstruct %int48, %int512, %int64 : (!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.aten.transpose.int %arg298, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %347 = torch.aten.bmm %346, %345 : !torch.tensor, !torch.tensor -> !torch.tensor
    %348 = torch.aten.transpose.int %arg299, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %349 = torch.aten.bmm %345, %348 : !torch.tensor, !torch.tensor -> !torch.tensor
    %350 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %351 = torch.aten.view %347, %350 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %352 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %353 = torch.aten.view %349, %352 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %354 = torch.aten.mul.Tensor %353, %arg297 : !torch.tensor, !torch.tensor -> !torch.tensor
    %355 = torch.aten._softmax_backward_data %354, %arg296, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %356 = torch.aten.div.Tensor %355, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %357 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %358 = torch.aten.view %356, %357 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %359 = torch.aten.transpose.int %arg294, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %360 = torch.aten.bmm %359, %358 : !torch.tensor, !torch.tensor -> !torch.tensor
    %361 = torch.aten.transpose.int %arg295, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %362 = torch.aten.bmm %358, %361 : !torch.tensor, !torch.tensor -> !torch.tensor
    %363 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %364 = torch.aten.view %360, %363 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %365 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %366 = torch.aten.view %362, %365 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %367 = torch.aten.transpose.int %364, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %368 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %369 = torch.aten.permute %366, %368 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %370 = torch.aten.clone %369, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %371 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %372 = torch.aten._unsafe_view %370, %371 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %373 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %374 = torch.aten.permute %351, %373 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %375 = torch.aten.clone %374, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %376 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %377 = torch.aten._unsafe_view %375, %376 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %378 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %379 = torch.aten.view %377, %378 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %380 = torch.aten.t %arg292 : !torch.tensor -> !torch.tensor
    %381 = torch.aten.mm %379, %380 : !torch.tensor, !torch.tensor -> !torch.tensor
    %382 = torch.aten.t %379 : !torch.tensor -> !torch.tensor
    %383 = torch.aten.mm %382, %arg293 : !torch.tensor, !torch.tensor -> !torch.tensor
    %384 = torch.aten.t %383 : !torch.tensor -> !torch.tensor
    %385 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %386 = torch.aten.sum.dim_IntList %379, %385, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %387 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %388 = torch.aten.view %386, %387 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %389 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %390 = torch.aten.view %381, %389 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %391 = torch.aten.add.Tensor %318, %390, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %392 = torch.aten.t %384 : !torch.tensor -> !torch.tensor
    %393 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %394 = torch.aten.permute %367, %393 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %395 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %396 = torch.aten.view %394, %395 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %397 = torch.aten.clone %396, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %398 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %399 = torch.aten._unsafe_view %397, %398 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %400 = torch.aten.t %arg290 : !torch.tensor -> !torch.tensor
    %401 = torch.aten.mm %399, %400 : !torch.tensor, !torch.tensor -> !torch.tensor
    %402 = torch.aten.t %399 : !torch.tensor -> !torch.tensor
    %403 = torch.aten.mm %402, %arg291 : !torch.tensor, !torch.tensor -> !torch.tensor
    %404 = torch.aten.t %403 : !torch.tensor -> !torch.tensor
    %405 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %406 = torch.aten.sum.dim_IntList %399, %405, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %407 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %408 = torch.aten.view %406, %407 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %409 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %410 = torch.aten.view %401, %409 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %411 = torch.aten.add.Tensor %391, %410, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %412 = torch.aten.t %404 : !torch.tensor -> !torch.tensor
    %413 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %414 = torch.aten.view %372, %413 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %415 = torch.aten.t %arg288 : !torch.tensor -> !torch.tensor
    %416 = torch.aten.mm %414, %415 : !torch.tensor, !torch.tensor -> !torch.tensor
    %417 = torch.aten.t %414 : !torch.tensor -> !torch.tensor
    %418 = torch.aten.mm %417, %arg289 : !torch.tensor, !torch.tensor -> !torch.tensor
    %419 = torch.aten.t %418 : !torch.tensor -> !torch.tensor
    %420 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %421 = torch.aten.sum.dim_IntList %414, %420, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %422 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %423 = torch.aten.view %421, %422 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %424 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %425 = torch.aten.view %416, %424 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %426 = torch.aten.add.Tensor %411, %425, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %427 = torch.aten.t %419 : !torch.tensor -> !torch.tensor
    %428 = torch.aten.sub.Tensor %arg287, %result1_58, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %429 = torch.aten.mul.Tensor %428, %result2_59 : !torch.tensor, !torch.tensor -> !torch.tensor
    %430 = torch.aten.mul.Tensor %426, %arg50 : !torch.tensor, !torch.tensor -> !torch.tensor
    %431 = torch.aten.mul.Tensor %430, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %432 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %433 = torch.aten.sum.dim_IntList %430, %432, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %434 = torch.aten.mul.Tensor %430, %429 : !torch.tensor, !torch.tensor -> !torch.tensor
    %435 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %436 = torch.aten.sum.dim_IntList %434, %435, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %437 = torch.aten.mul.Tensor %429, %436 : !torch.tensor, !torch.tensor -> !torch.tensor
    %438 = torch.aten.sub.Tensor %431, %433, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %439 = torch.aten.sub.Tensor %438, %437, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %440 = torch.aten.div.Tensor %result2_59, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %441 = torch.aten.mul.Tensor %440, %439 : !torch.tensor, !torch.tensor -> !torch.tensor
    %442 = torch.aten.mul.Tensor %426, %429 : !torch.tensor, !torch.tensor -> !torch.tensor
    %443 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %444 = torch.aten.sum.dim_IntList %442, %443, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %445 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %446 = torch.aten.sum.dim_IntList %426, %445, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %447 = torch.aten.mul.Tensor %441, %arg286 : !torch.tensor, !torch.tensor -> !torch.tensor
    %448 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %449 = torch.aten.view %447, %448 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %450 = torch.aten.t %arg284 : !torch.tensor -> !torch.tensor
    %451 = torch.aten.mm %449, %450 : !torch.tensor, !torch.tensor -> !torch.tensor
    %452 = torch.aten.t %449 : !torch.tensor -> !torch.tensor
    %453 = torch.aten.mm %452, %arg285 : !torch.tensor, !torch.tensor -> !torch.tensor
    %454 = torch.aten.t %453 : !torch.tensor -> !torch.tensor
    %455 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %456 = torch.aten.sum.dim_IntList %449, %455, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %457 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %458 = torch.aten.view %456, %457 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %459 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %460 = torch.aten.view %451, %459 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %461 = torch.aten.t %454 : !torch.tensor -> !torch.tensor
    %462 = torch.aten.gelu_backward %460, %arg283, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %463 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %464 = torch.aten.view %462, %463 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %465 = torch.aten.t %arg281 : !torch.tensor -> !torch.tensor
    %466 = torch.aten.mm %464, %465 : !torch.tensor, !torch.tensor -> !torch.tensor
    %467 = torch.aten.t %464 : !torch.tensor -> !torch.tensor
    %468 = torch.aten.mm %467, %arg282 : !torch.tensor, !torch.tensor -> !torch.tensor
    %469 = torch.aten.t %468 : !torch.tensor -> !torch.tensor
    %470 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %471 = torch.aten.sum.dim_IntList %464, %470, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %472 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %473 = torch.aten.view %471, %472 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %474 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %475 = torch.aten.view %466, %474 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %476 = torch.aten.add.Tensor %441, %475, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %477 = torch.aten.t %469 : !torch.tensor -> !torch.tensor
    %478 = torch.aten.sub.Tensor %arg280, %result1_55, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %479 = torch.aten.mul.Tensor %478, %result2_56 : !torch.tensor, !torch.tensor -> !torch.tensor
    %480 = torch.aten.mul.Tensor %476, %arg48 : !torch.tensor, !torch.tensor -> !torch.tensor
    %481 = torch.aten.mul.Tensor %480, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %482 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %483 = torch.aten.sum.dim_IntList %480, %482, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %484 = torch.aten.mul.Tensor %480, %479 : !torch.tensor, !torch.tensor -> !torch.tensor
    %485 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %486 = torch.aten.sum.dim_IntList %484, %485, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %487 = torch.aten.mul.Tensor %479, %486 : !torch.tensor, !torch.tensor -> !torch.tensor
    %488 = torch.aten.sub.Tensor %481, %483, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %489 = torch.aten.sub.Tensor %488, %487, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %490 = torch.aten.div.Tensor %result2_56, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %491 = torch.aten.mul.Tensor %490, %489 : !torch.tensor, !torch.tensor -> !torch.tensor
    %492 = torch.aten.mul.Tensor %476, %479 : !torch.tensor, !torch.tensor -> !torch.tensor
    %493 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %494 = torch.aten.sum.dim_IntList %492, %493, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %495 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %496 = torch.aten.sum.dim_IntList %476, %495, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %497 = torch.aten.mul.Tensor %491, %arg279 : !torch.tensor, !torch.tensor -> !torch.tensor
    %498 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %499 = torch.aten.view %497, %498 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %500 = torch.aten.t %arg277 : !torch.tensor -> !torch.tensor
    %501 = torch.aten.mm %499, %500 : !torch.tensor, !torch.tensor -> !torch.tensor
    %502 = torch.aten.t %499 : !torch.tensor -> !torch.tensor
    %503 = torch.aten.mm %502, %arg278 : !torch.tensor, !torch.tensor -> !torch.tensor
    %504 = torch.aten.t %503 : !torch.tensor -> !torch.tensor
    %505 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %506 = torch.aten.sum.dim_IntList %499, %505, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %507 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %508 = torch.aten.view %506, %507 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %509 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %510 = torch.aten.view %501, %509 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %511 = torch.aten.t %504 : !torch.tensor -> !torch.tensor
    %512 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %513 = torch.aten.view %510, %512 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %514 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %515 = torch.aten.permute %513, %514 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %516 = torch.aten.clone %515, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %517 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %518 = torch.aten._unsafe_view %516, %517 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %519 = torch.aten.transpose.int %arg275, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %520 = torch.aten.bmm %519, %518 : !torch.tensor, !torch.tensor -> !torch.tensor
    %521 = torch.aten.transpose.int %arg276, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %522 = torch.aten.bmm %518, %521 : !torch.tensor, !torch.tensor -> !torch.tensor
    %523 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %524 = torch.aten.view %520, %523 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %525 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %526 = torch.aten.view %522, %525 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %527 = torch.aten.mul.Tensor %526, %arg274 : !torch.tensor, !torch.tensor -> !torch.tensor
    %528 = torch.aten._softmax_backward_data %527, %arg273, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %529 = torch.aten.div.Tensor %528, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %530 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %531 = torch.aten.view %529, %530 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %532 = torch.aten.transpose.int %arg271, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %533 = torch.aten.bmm %532, %531 : !torch.tensor, !torch.tensor -> !torch.tensor
    %534 = torch.aten.transpose.int %arg272, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %535 = torch.aten.bmm %531, %534 : !torch.tensor, !torch.tensor -> !torch.tensor
    %536 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %537 = torch.aten.view %533, %536 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %538 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %539 = torch.aten.view %535, %538 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %540 = torch.aten.transpose.int %537, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %541 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %542 = torch.aten.permute %539, %541 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %543 = torch.aten.clone %542, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %544 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %545 = torch.aten._unsafe_view %543, %544 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %546 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %547 = torch.aten.permute %524, %546 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %548 = torch.aten.clone %547, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %549 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %550 = torch.aten._unsafe_view %548, %549 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %551 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %552 = torch.aten.view %550, %551 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %553 = torch.aten.t %arg269 : !torch.tensor -> !torch.tensor
    %554 = torch.aten.mm %552, %553 : !torch.tensor, !torch.tensor -> !torch.tensor
    %555 = torch.aten.t %552 : !torch.tensor -> !torch.tensor
    %556 = torch.aten.mm %555, %arg270 : !torch.tensor, !torch.tensor -> !torch.tensor
    %557 = torch.aten.t %556 : !torch.tensor -> !torch.tensor
    %558 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %559 = torch.aten.sum.dim_IntList %552, %558, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %560 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %561 = torch.aten.view %559, %560 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %562 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %563 = torch.aten.view %554, %562 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %564 = torch.aten.add.Tensor %491, %563, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %565 = torch.aten.t %557 : !torch.tensor -> !torch.tensor
    %566 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %567 = torch.aten.permute %540, %566 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %568 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %569 = torch.aten.view %567, %568 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %570 = torch.aten.clone %569, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %571 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %572 = torch.aten._unsafe_view %570, %571 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %573 = torch.aten.t %arg267 : !torch.tensor -> !torch.tensor
    %574 = torch.aten.mm %572, %573 : !torch.tensor, !torch.tensor -> !torch.tensor
    %575 = torch.aten.t %572 : !torch.tensor -> !torch.tensor
    %576 = torch.aten.mm %575, %arg268 : !torch.tensor, !torch.tensor -> !torch.tensor
    %577 = torch.aten.t %576 : !torch.tensor -> !torch.tensor
    %578 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %579 = torch.aten.sum.dim_IntList %572, %578, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %580 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %581 = torch.aten.view %579, %580 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %582 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %583 = torch.aten.view %574, %582 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %584 = torch.aten.add.Tensor %564, %583, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %585 = torch.aten.t %577 : !torch.tensor -> !torch.tensor
    %586 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %587 = torch.aten.view %545, %586 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %588 = torch.aten.t %arg265 : !torch.tensor -> !torch.tensor
    %589 = torch.aten.mm %587, %588 : !torch.tensor, !torch.tensor -> !torch.tensor
    %590 = torch.aten.t %587 : !torch.tensor -> !torch.tensor
    %591 = torch.aten.mm %590, %arg266 : !torch.tensor, !torch.tensor -> !torch.tensor
    %592 = torch.aten.t %591 : !torch.tensor -> !torch.tensor
    %593 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %594 = torch.aten.sum.dim_IntList %587, %593, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %595 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %596 = torch.aten.view %594, %595 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %597 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %598 = torch.aten.view %589, %597 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %599 = torch.aten.add.Tensor %584, %598, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %600 = torch.aten.t %592 : !torch.tensor -> !torch.tensor
    %601 = torch.aten.sub.Tensor %arg264, %result1_52, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %602 = torch.aten.mul.Tensor %601, %result2_53 : !torch.tensor, !torch.tensor -> !torch.tensor
    %603 = torch.aten.mul.Tensor %599, %arg46 : !torch.tensor, !torch.tensor -> !torch.tensor
    %604 = torch.aten.mul.Tensor %603, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %605 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %606 = torch.aten.sum.dim_IntList %603, %605, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %607 = torch.aten.mul.Tensor %603, %602 : !torch.tensor, !torch.tensor -> !torch.tensor
    %608 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %609 = torch.aten.sum.dim_IntList %607, %608, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %610 = torch.aten.mul.Tensor %602, %609 : !torch.tensor, !torch.tensor -> !torch.tensor
    %611 = torch.aten.sub.Tensor %604, %606, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %612 = torch.aten.sub.Tensor %611, %610, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %613 = torch.aten.div.Tensor %result2_53, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %614 = torch.aten.mul.Tensor %613, %612 : !torch.tensor, !torch.tensor -> !torch.tensor
    %615 = torch.aten.mul.Tensor %599, %602 : !torch.tensor, !torch.tensor -> !torch.tensor
    %616 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %617 = torch.aten.sum.dim_IntList %615, %616, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %618 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %619 = torch.aten.sum.dim_IntList %599, %618, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %620 = torch.aten.mul.Tensor %614, %arg263 : !torch.tensor, !torch.tensor -> !torch.tensor
    %621 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %622 = torch.aten.view %620, %621 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %623 = torch.aten.t %arg261 : !torch.tensor -> !torch.tensor
    %624 = torch.aten.mm %622, %623 : !torch.tensor, !torch.tensor -> !torch.tensor
    %625 = torch.aten.t %622 : !torch.tensor -> !torch.tensor
    %626 = torch.aten.mm %625, %arg262 : !torch.tensor, !torch.tensor -> !torch.tensor
    %627 = torch.aten.t %626 : !torch.tensor -> !torch.tensor
    %628 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %629 = torch.aten.sum.dim_IntList %622, %628, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %630 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %631 = torch.aten.view %629, %630 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %632 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %633 = torch.aten.view %624, %632 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %634 = torch.aten.t %627 : !torch.tensor -> !torch.tensor
    %635 = torch.aten.gelu_backward %633, %arg260, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %636 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %637 = torch.aten.view %635, %636 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %638 = torch.aten.t %arg258 : !torch.tensor -> !torch.tensor
    %639 = torch.aten.mm %637, %638 : !torch.tensor, !torch.tensor -> !torch.tensor
    %640 = torch.aten.t %637 : !torch.tensor -> !torch.tensor
    %641 = torch.aten.mm %640, %arg259 : !torch.tensor, !torch.tensor -> !torch.tensor
    %642 = torch.aten.t %641 : !torch.tensor -> !torch.tensor
    %643 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %644 = torch.aten.sum.dim_IntList %637, %643, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %645 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %646 = torch.aten.view %644, %645 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %647 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %648 = torch.aten.view %639, %647 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %649 = torch.aten.add.Tensor %614, %648, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %650 = torch.aten.t %642 : !torch.tensor -> !torch.tensor
    %651 = torch.aten.sub.Tensor %arg257, %result1_49, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %652 = torch.aten.mul.Tensor %651, %result2_50 : !torch.tensor, !torch.tensor -> !torch.tensor
    %653 = torch.aten.mul.Tensor %649, %arg44 : !torch.tensor, !torch.tensor -> !torch.tensor
    %654 = torch.aten.mul.Tensor %653, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %655 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %656 = torch.aten.sum.dim_IntList %653, %655, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %657 = torch.aten.mul.Tensor %653, %652 : !torch.tensor, !torch.tensor -> !torch.tensor
    %658 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %659 = torch.aten.sum.dim_IntList %657, %658, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %660 = torch.aten.mul.Tensor %652, %659 : !torch.tensor, !torch.tensor -> !torch.tensor
    %661 = torch.aten.sub.Tensor %654, %656, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %662 = torch.aten.sub.Tensor %661, %660, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %663 = torch.aten.div.Tensor %result2_50, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %664 = torch.aten.mul.Tensor %663, %662 : !torch.tensor, !torch.tensor -> !torch.tensor
    %665 = torch.aten.mul.Tensor %649, %652 : !torch.tensor, !torch.tensor -> !torch.tensor
    %666 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %667 = torch.aten.sum.dim_IntList %665, %666, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %668 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %669 = torch.aten.sum.dim_IntList %649, %668, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %670 = torch.aten.mul.Tensor %664, %arg256 : !torch.tensor, !torch.tensor -> !torch.tensor
    %671 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %672 = torch.aten.view %670, %671 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %673 = torch.aten.t %arg254 : !torch.tensor -> !torch.tensor
    %674 = torch.aten.mm %672, %673 : !torch.tensor, !torch.tensor -> !torch.tensor
    %675 = torch.aten.t %672 : !torch.tensor -> !torch.tensor
    %676 = torch.aten.mm %675, %arg255 : !torch.tensor, !torch.tensor -> !torch.tensor
    %677 = torch.aten.t %676 : !torch.tensor -> !torch.tensor
    %678 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %679 = torch.aten.sum.dim_IntList %672, %678, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %680 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %681 = torch.aten.view %679, %680 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %682 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %683 = torch.aten.view %674, %682 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %684 = torch.aten.t %677 : !torch.tensor -> !torch.tensor
    %685 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %686 = torch.aten.view %683, %685 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %687 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %688 = torch.aten.permute %686, %687 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %689 = torch.aten.clone %688, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %690 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %691 = torch.aten._unsafe_view %689, %690 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %692 = torch.aten.transpose.int %arg252, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %693 = torch.aten.bmm %692, %691 : !torch.tensor, !torch.tensor -> !torch.tensor
    %694 = torch.aten.transpose.int %arg253, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %695 = torch.aten.bmm %691, %694 : !torch.tensor, !torch.tensor -> !torch.tensor
    %696 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %697 = torch.aten.view %693, %696 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %698 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %699 = torch.aten.view %695, %698 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %700 = torch.aten.mul.Tensor %699, %arg251 : !torch.tensor, !torch.tensor -> !torch.tensor
    %701 = torch.aten._softmax_backward_data %700, %arg250, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %702 = torch.aten.div.Tensor %701, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %703 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %704 = torch.aten.view %702, %703 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %705 = torch.aten.transpose.int %arg248, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %706 = torch.aten.bmm %705, %704 : !torch.tensor, !torch.tensor -> !torch.tensor
    %707 = torch.aten.transpose.int %arg249, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %708 = torch.aten.bmm %704, %707 : !torch.tensor, !torch.tensor -> !torch.tensor
    %709 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %710 = torch.aten.view %706, %709 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %711 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %712 = torch.aten.view %708, %711 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %713 = torch.aten.transpose.int %710, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %714 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %715 = torch.aten.permute %712, %714 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %716 = torch.aten.clone %715, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %717 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %718 = torch.aten._unsafe_view %716, %717 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %719 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %720 = torch.aten.permute %697, %719 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %721 = torch.aten.clone %720, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %722 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %723 = torch.aten._unsafe_view %721, %722 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %724 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %725 = torch.aten.view %723, %724 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %726 = torch.aten.t %arg246 : !torch.tensor -> !torch.tensor
    %727 = torch.aten.mm %725, %726 : !torch.tensor, !torch.tensor -> !torch.tensor
    %728 = torch.aten.t %725 : !torch.tensor -> !torch.tensor
    %729 = torch.aten.mm %728, %arg247 : !torch.tensor, !torch.tensor -> !torch.tensor
    %730 = torch.aten.t %729 : !torch.tensor -> !torch.tensor
    %731 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %732 = torch.aten.sum.dim_IntList %725, %731, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %733 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %734 = torch.aten.view %732, %733 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %735 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %736 = torch.aten.view %727, %735 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %737 = torch.aten.add.Tensor %664, %736, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %738 = torch.aten.t %730 : !torch.tensor -> !torch.tensor
    %739 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %740 = torch.aten.permute %713, %739 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %741 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %742 = torch.aten.view %740, %741 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %743 = torch.aten.clone %742, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %744 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %745 = torch.aten._unsafe_view %743, %744 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %746 = torch.aten.t %arg244 : !torch.tensor -> !torch.tensor
    %747 = torch.aten.mm %745, %746 : !torch.tensor, !torch.tensor -> !torch.tensor
    %748 = torch.aten.t %745 : !torch.tensor -> !torch.tensor
    %749 = torch.aten.mm %748, %arg245 : !torch.tensor, !torch.tensor -> !torch.tensor
    %750 = torch.aten.t %749 : !torch.tensor -> !torch.tensor
    %751 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %752 = torch.aten.sum.dim_IntList %745, %751, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %753 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %754 = torch.aten.view %752, %753 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %755 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %756 = torch.aten.view %747, %755 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %757 = torch.aten.add.Tensor %737, %756, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %758 = torch.aten.t %750 : !torch.tensor -> !torch.tensor
    %759 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %760 = torch.aten.view %718, %759 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %761 = torch.aten.t %arg242 : !torch.tensor -> !torch.tensor
    %762 = torch.aten.mm %760, %761 : !torch.tensor, !torch.tensor -> !torch.tensor
    %763 = torch.aten.t %760 : !torch.tensor -> !torch.tensor
    %764 = torch.aten.mm %763, %arg243 : !torch.tensor, !torch.tensor -> !torch.tensor
    %765 = torch.aten.t %764 : !torch.tensor -> !torch.tensor
    %766 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %767 = torch.aten.sum.dim_IntList %760, %766, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %768 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %769 = torch.aten.view %767, %768 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %770 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %771 = torch.aten.view %762, %770 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %772 = torch.aten.add.Tensor %757, %771, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %773 = torch.aten.t %765 : !torch.tensor -> !torch.tensor
    %774 = torch.aten.sub.Tensor %arg241, %result1_46, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %775 = torch.aten.mul.Tensor %774, %result2_47 : !torch.tensor, !torch.tensor -> !torch.tensor
    %776 = torch.aten.mul.Tensor %772, %arg42 : !torch.tensor, !torch.tensor -> !torch.tensor
    %777 = torch.aten.mul.Tensor %776, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %778 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %779 = torch.aten.sum.dim_IntList %776, %778, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %780 = torch.aten.mul.Tensor %776, %775 : !torch.tensor, !torch.tensor -> !torch.tensor
    %781 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %782 = torch.aten.sum.dim_IntList %780, %781, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %783 = torch.aten.mul.Tensor %775, %782 : !torch.tensor, !torch.tensor -> !torch.tensor
    %784 = torch.aten.sub.Tensor %777, %779, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %785 = torch.aten.sub.Tensor %784, %783, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %786 = torch.aten.div.Tensor %result2_47, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %787 = torch.aten.mul.Tensor %786, %785 : !torch.tensor, !torch.tensor -> !torch.tensor
    %788 = torch.aten.mul.Tensor %772, %775 : !torch.tensor, !torch.tensor -> !torch.tensor
    %789 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %790 = torch.aten.sum.dim_IntList %788, %789, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %791 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %792 = torch.aten.sum.dim_IntList %772, %791, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %793 = torch.aten.mul.Tensor %787, %arg240 : !torch.tensor, !torch.tensor -> !torch.tensor
    %794 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %795 = torch.aten.view %793, %794 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %796 = torch.aten.t %arg238 : !torch.tensor -> !torch.tensor
    %797 = torch.aten.mm %795, %796 : !torch.tensor, !torch.tensor -> !torch.tensor
    %798 = torch.aten.t %795 : !torch.tensor -> !torch.tensor
    %799 = torch.aten.mm %798, %arg239 : !torch.tensor, !torch.tensor -> !torch.tensor
    %800 = torch.aten.t %799 : !torch.tensor -> !torch.tensor
    %801 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %802 = torch.aten.sum.dim_IntList %795, %801, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %803 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %804 = torch.aten.view %802, %803 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %805 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %806 = torch.aten.view %797, %805 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %807 = torch.aten.t %800 : !torch.tensor -> !torch.tensor
    %808 = torch.aten.gelu_backward %806, %arg237, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %809 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %810 = torch.aten.view %808, %809 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %811 = torch.aten.t %arg235 : !torch.tensor -> !torch.tensor
    %812 = torch.aten.mm %810, %811 : !torch.tensor, !torch.tensor -> !torch.tensor
    %813 = torch.aten.t %810 : !torch.tensor -> !torch.tensor
    %814 = torch.aten.mm %813, %arg236 : !torch.tensor, !torch.tensor -> !torch.tensor
    %815 = torch.aten.t %814 : !torch.tensor -> !torch.tensor
    %816 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %817 = torch.aten.sum.dim_IntList %810, %816, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %818 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %819 = torch.aten.view %817, %818 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %820 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %821 = torch.aten.view %812, %820 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %822 = torch.aten.add.Tensor %787, %821, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %823 = torch.aten.t %815 : !torch.tensor -> !torch.tensor
    %824 = torch.aten.sub.Tensor %arg234, %result1_43, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %825 = torch.aten.mul.Tensor %824, %result2_44 : !torch.tensor, !torch.tensor -> !torch.tensor
    %826 = torch.aten.mul.Tensor %822, %arg40 : !torch.tensor, !torch.tensor -> !torch.tensor
    %827 = torch.aten.mul.Tensor %826, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %828 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %829 = torch.aten.sum.dim_IntList %826, %828, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %830 = torch.aten.mul.Tensor %826, %825 : !torch.tensor, !torch.tensor -> !torch.tensor
    %831 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %832 = torch.aten.sum.dim_IntList %830, %831, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %833 = torch.aten.mul.Tensor %825, %832 : !torch.tensor, !torch.tensor -> !torch.tensor
    %834 = torch.aten.sub.Tensor %827, %829, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %835 = torch.aten.sub.Tensor %834, %833, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %836 = torch.aten.div.Tensor %result2_44, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %837 = torch.aten.mul.Tensor %836, %835 : !torch.tensor, !torch.tensor -> !torch.tensor
    %838 = torch.aten.mul.Tensor %822, %825 : !torch.tensor, !torch.tensor -> !torch.tensor
    %839 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %840 = torch.aten.sum.dim_IntList %838, %839, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %841 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %842 = torch.aten.sum.dim_IntList %822, %841, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %843 = torch.aten.mul.Tensor %837, %arg233 : !torch.tensor, !torch.tensor -> !torch.tensor
    %844 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %845 = torch.aten.view %843, %844 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %846 = torch.aten.t %arg231 : !torch.tensor -> !torch.tensor
    %847 = torch.aten.mm %845, %846 : !torch.tensor, !torch.tensor -> !torch.tensor
    %848 = torch.aten.t %845 : !torch.tensor -> !torch.tensor
    %849 = torch.aten.mm %848, %arg232 : !torch.tensor, !torch.tensor -> !torch.tensor
    %850 = torch.aten.t %849 : !torch.tensor -> !torch.tensor
    %851 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %852 = torch.aten.sum.dim_IntList %845, %851, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %853 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %854 = torch.aten.view %852, %853 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %855 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %856 = torch.aten.view %847, %855 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %857 = torch.aten.t %850 : !torch.tensor -> !torch.tensor
    %858 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %859 = torch.aten.view %856, %858 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %860 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %861 = torch.aten.permute %859, %860 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %862 = torch.aten.clone %861, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %863 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %864 = torch.aten._unsafe_view %862, %863 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %865 = torch.aten.transpose.int %arg229, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %866 = torch.aten.bmm %865, %864 : !torch.tensor, !torch.tensor -> !torch.tensor
    %867 = torch.aten.transpose.int %arg230, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %868 = torch.aten.bmm %864, %867 : !torch.tensor, !torch.tensor -> !torch.tensor
    %869 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %870 = torch.aten.view %866, %869 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %871 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %872 = torch.aten.view %868, %871 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %873 = torch.aten.mul.Tensor %872, %arg228 : !torch.tensor, !torch.tensor -> !torch.tensor
    %874 = torch.aten._softmax_backward_data %873, %arg227, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %875 = torch.aten.div.Tensor %874, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %876 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %877 = torch.aten.view %875, %876 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %878 = torch.aten.transpose.int %arg225, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %879 = torch.aten.bmm %878, %877 : !torch.tensor, !torch.tensor -> !torch.tensor
    %880 = torch.aten.transpose.int %arg226, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %881 = torch.aten.bmm %877, %880 : !torch.tensor, !torch.tensor -> !torch.tensor
    %882 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %883 = torch.aten.view %879, %882 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %884 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %885 = torch.aten.view %881, %884 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %886 = torch.aten.transpose.int %883, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %887 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %888 = torch.aten.permute %885, %887 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %889 = torch.aten.clone %888, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %890 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %891 = torch.aten._unsafe_view %889, %890 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %892 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %893 = torch.aten.permute %870, %892 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %894 = torch.aten.clone %893, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %895 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %896 = torch.aten._unsafe_view %894, %895 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %897 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %898 = torch.aten.view %896, %897 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %899 = torch.aten.t %arg223 : !torch.tensor -> !torch.tensor
    %900 = torch.aten.mm %898, %899 : !torch.tensor, !torch.tensor -> !torch.tensor
    %901 = torch.aten.t %898 : !torch.tensor -> !torch.tensor
    %902 = torch.aten.mm %901, %arg224 : !torch.tensor, !torch.tensor -> !torch.tensor
    %903 = torch.aten.t %902 : !torch.tensor -> !torch.tensor
    %904 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %905 = torch.aten.sum.dim_IntList %898, %904, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %906 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %907 = torch.aten.view %905, %906 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %908 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %909 = torch.aten.view %900, %908 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %910 = torch.aten.add.Tensor %837, %909, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %911 = torch.aten.t %903 : !torch.tensor -> !torch.tensor
    %912 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %913 = torch.aten.permute %886, %912 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %914 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %915 = torch.aten.view %913, %914 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %916 = torch.aten.clone %915, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %917 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %918 = torch.aten._unsafe_view %916, %917 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %919 = torch.aten.t %arg221 : !torch.tensor -> !torch.tensor
    %920 = torch.aten.mm %918, %919 : !torch.tensor, !torch.tensor -> !torch.tensor
    %921 = torch.aten.t %918 : !torch.tensor -> !torch.tensor
    %922 = torch.aten.mm %921, %arg222 : !torch.tensor, !torch.tensor -> !torch.tensor
    %923 = torch.aten.t %922 : !torch.tensor -> !torch.tensor
    %924 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %925 = torch.aten.sum.dim_IntList %918, %924, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %926 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %927 = torch.aten.view %925, %926 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %928 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %929 = torch.aten.view %920, %928 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %930 = torch.aten.add.Tensor %910, %929, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %931 = torch.aten.t %923 : !torch.tensor -> !torch.tensor
    %932 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %933 = torch.aten.view %891, %932 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %934 = torch.aten.t %arg219 : !torch.tensor -> !torch.tensor
    %935 = torch.aten.mm %933, %934 : !torch.tensor, !torch.tensor -> !torch.tensor
    %936 = torch.aten.t %933 : !torch.tensor -> !torch.tensor
    %937 = torch.aten.mm %936, %arg220 : !torch.tensor, !torch.tensor -> !torch.tensor
    %938 = torch.aten.t %937 : !torch.tensor -> !torch.tensor
    %939 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %940 = torch.aten.sum.dim_IntList %933, %939, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %941 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %942 = torch.aten.view %940, %941 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %943 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %944 = torch.aten.view %935, %943 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %945 = torch.aten.add.Tensor %930, %944, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %946 = torch.aten.t %938 : !torch.tensor -> !torch.tensor
    %947 = torch.aten.sub.Tensor %arg218, %result1_40, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %948 = torch.aten.mul.Tensor %947, %result2_41 : !torch.tensor, !torch.tensor -> !torch.tensor
    %949 = torch.aten.mul.Tensor %945, %arg38 : !torch.tensor, !torch.tensor -> !torch.tensor
    %950 = torch.aten.mul.Tensor %949, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %951 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %952 = torch.aten.sum.dim_IntList %949, %951, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %953 = torch.aten.mul.Tensor %949, %948 : !torch.tensor, !torch.tensor -> !torch.tensor
    %954 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %955 = torch.aten.sum.dim_IntList %953, %954, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %956 = torch.aten.mul.Tensor %948, %955 : !torch.tensor, !torch.tensor -> !torch.tensor
    %957 = torch.aten.sub.Tensor %950, %952, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %958 = torch.aten.sub.Tensor %957, %956, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %959 = torch.aten.div.Tensor %result2_41, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %960 = torch.aten.mul.Tensor %959, %958 : !torch.tensor, !torch.tensor -> !torch.tensor
    %961 = torch.aten.mul.Tensor %945, %948 : !torch.tensor, !torch.tensor -> !torch.tensor
    %962 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %963 = torch.aten.sum.dim_IntList %961, %962, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %964 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %965 = torch.aten.sum.dim_IntList %945, %964, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %966 = torch.aten.mul.Tensor %960, %arg217 : !torch.tensor, !torch.tensor -> !torch.tensor
    %967 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %968 = torch.aten.view %966, %967 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %969 = torch.aten.t %arg215 : !torch.tensor -> !torch.tensor
    %970 = torch.aten.mm %968, %969 : !torch.tensor, !torch.tensor -> !torch.tensor
    %971 = torch.aten.t %968 : !torch.tensor -> !torch.tensor
    %972 = torch.aten.mm %971, %arg216 : !torch.tensor, !torch.tensor -> !torch.tensor
    %973 = torch.aten.t %972 : !torch.tensor -> !torch.tensor
    %974 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %975 = torch.aten.sum.dim_IntList %968, %974, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %976 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %977 = torch.aten.view %975, %976 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %978 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %979 = torch.aten.view %970, %978 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %980 = torch.aten.t %973 : !torch.tensor -> !torch.tensor
    %981 = torch.aten.gelu_backward %979, %arg214, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %982 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %983 = torch.aten.view %981, %982 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %984 = torch.aten.t %arg212 : !torch.tensor -> !torch.tensor
    %985 = torch.aten.mm %983, %984 : !torch.tensor, !torch.tensor -> !torch.tensor
    %986 = torch.aten.t %983 : !torch.tensor -> !torch.tensor
    %987 = torch.aten.mm %986, %arg213 : !torch.tensor, !torch.tensor -> !torch.tensor
    %988 = torch.aten.t %987 : !torch.tensor -> !torch.tensor
    %989 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %990 = torch.aten.sum.dim_IntList %983, %989, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %991 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %992 = torch.aten.view %990, %991 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %993 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %994 = torch.aten.view %985, %993 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %995 = torch.aten.add.Tensor %960, %994, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %996 = torch.aten.t %988 : !torch.tensor -> !torch.tensor
    %997 = torch.aten.sub.Tensor %arg211, %result1_37, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %998 = torch.aten.mul.Tensor %997, %result2_38 : !torch.tensor, !torch.tensor -> !torch.tensor
    %999 = torch.aten.mul.Tensor %995, %arg36 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1000 = torch.aten.mul.Tensor %999, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1001 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1002 = torch.aten.sum.dim_IntList %999, %1001, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1003 = torch.aten.mul.Tensor %999, %998 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1004 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1005 = torch.aten.sum.dim_IntList %1003, %1004, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1006 = torch.aten.mul.Tensor %998, %1005 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1007 = torch.aten.sub.Tensor %1000, %1002, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1008 = torch.aten.sub.Tensor %1007, %1006, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1009 = torch.aten.div.Tensor %result2_38, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1010 = torch.aten.mul.Tensor %1009, %1008 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1011 = torch.aten.mul.Tensor %995, %998 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1012 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1013 = torch.aten.sum.dim_IntList %1011, %1012, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1014 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1015 = torch.aten.sum.dim_IntList %995, %1014, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1016 = torch.aten.mul.Tensor %1010, %arg210 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1017 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1018 = torch.aten.view %1016, %1017 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1019 = torch.aten.t %arg208 : !torch.tensor -> !torch.tensor
    %1020 = torch.aten.mm %1018, %1019 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1021 = torch.aten.t %1018 : !torch.tensor -> !torch.tensor
    %1022 = torch.aten.mm %1021, %arg209 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1023 = torch.aten.t %1022 : !torch.tensor -> !torch.tensor
    %1024 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1025 = torch.aten.sum.dim_IntList %1018, %1024, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1026 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1027 = torch.aten.view %1025, %1026 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1028 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1029 = torch.aten.view %1020, %1028 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1030 = torch.aten.t %1023 : !torch.tensor -> !torch.tensor
    %1031 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1032 = torch.aten.view %1029, %1031 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1033 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1034 = torch.aten.permute %1032, %1033 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1035 = torch.aten.clone %1034, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1036 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1037 = torch.aten._unsafe_view %1035, %1036 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1038 = torch.aten.transpose.int %arg206, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1039 = torch.aten.bmm %1038, %1037 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1040 = torch.aten.transpose.int %arg207, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1041 = torch.aten.bmm %1037, %1040 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1042 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1043 = torch.aten.view %1039, %1042 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1044 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1045 = torch.aten.view %1041, %1044 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1046 = torch.aten.mul.Tensor %1045, %arg205 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1047 = torch.aten._softmax_backward_data %1046, %arg204, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1048 = torch.aten.div.Tensor %1047, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1049 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1050 = torch.aten.view %1048, %1049 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1051 = torch.aten.transpose.int %arg202, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1052 = torch.aten.bmm %1051, %1050 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1053 = torch.aten.transpose.int %arg203, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1054 = torch.aten.bmm %1050, %1053 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1055 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1056 = torch.aten.view %1052, %1055 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1057 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1058 = torch.aten.view %1054, %1057 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1059 = torch.aten.transpose.int %1056, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1060 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1061 = torch.aten.permute %1058, %1060 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1062 = torch.aten.clone %1061, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1063 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1064 = torch.aten._unsafe_view %1062, %1063 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1065 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1066 = torch.aten.permute %1043, %1065 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1067 = torch.aten.clone %1066, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1068 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1069 = torch.aten._unsafe_view %1067, %1068 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1070 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1071 = torch.aten.view %1069, %1070 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1072 = torch.aten.t %arg200 : !torch.tensor -> !torch.tensor
    %1073 = torch.aten.mm %1071, %1072 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1074 = torch.aten.t %1071 : !torch.tensor -> !torch.tensor
    %1075 = torch.aten.mm %1074, %arg201 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1076 = torch.aten.t %1075 : !torch.tensor -> !torch.tensor
    %1077 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1078 = torch.aten.sum.dim_IntList %1071, %1077, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1079 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1080 = torch.aten.view %1078, %1079 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1081 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1082 = torch.aten.view %1073, %1081 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1083 = torch.aten.add.Tensor %1010, %1082, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1084 = torch.aten.t %1076 : !torch.tensor -> !torch.tensor
    %1085 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1086 = torch.aten.permute %1059, %1085 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1087 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1088 = torch.aten.view %1086, %1087 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1089 = torch.aten.clone %1088, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1090 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1091 = torch.aten._unsafe_view %1089, %1090 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1092 = torch.aten.t %arg198 : !torch.tensor -> !torch.tensor
    %1093 = torch.aten.mm %1091, %1092 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1094 = torch.aten.t %1091 : !torch.tensor -> !torch.tensor
    %1095 = torch.aten.mm %1094, %arg199 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1096 = torch.aten.t %1095 : !torch.tensor -> !torch.tensor
    %1097 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1098 = torch.aten.sum.dim_IntList %1091, %1097, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1099 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1100 = torch.aten.view %1098, %1099 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1101 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1102 = torch.aten.view %1093, %1101 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1103 = torch.aten.add.Tensor %1083, %1102, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1104 = torch.aten.t %1096 : !torch.tensor -> !torch.tensor
    %1105 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1106 = torch.aten.view %1064, %1105 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1107 = torch.aten.t %arg196 : !torch.tensor -> !torch.tensor
    %1108 = torch.aten.mm %1106, %1107 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1109 = torch.aten.t %1106 : !torch.tensor -> !torch.tensor
    %1110 = torch.aten.mm %1109, %arg197 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1111 = torch.aten.t %1110 : !torch.tensor -> !torch.tensor
    %1112 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1113 = torch.aten.sum.dim_IntList %1106, %1112, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1114 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1115 = torch.aten.view %1113, %1114 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1116 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1117 = torch.aten.view %1108, %1116 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1118 = torch.aten.add.Tensor %1103, %1117, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1119 = torch.aten.t %1111 : !torch.tensor -> !torch.tensor
    %1120 = torch.aten.sub.Tensor %arg195, %result1_34, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1121 = torch.aten.mul.Tensor %1120, %result2_35 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1122 = torch.aten.mul.Tensor %1118, %arg34 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1123 = torch.aten.mul.Tensor %1122, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1124 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1125 = torch.aten.sum.dim_IntList %1122, %1124, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1126 = torch.aten.mul.Tensor %1122, %1121 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1127 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1128 = torch.aten.sum.dim_IntList %1126, %1127, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1129 = torch.aten.mul.Tensor %1121, %1128 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1130 = torch.aten.sub.Tensor %1123, %1125, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1131 = torch.aten.sub.Tensor %1130, %1129, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1132 = torch.aten.div.Tensor %result2_35, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1133 = torch.aten.mul.Tensor %1132, %1131 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1134 = torch.aten.mul.Tensor %1118, %1121 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1135 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1136 = torch.aten.sum.dim_IntList %1134, %1135, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1137 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1138 = torch.aten.sum.dim_IntList %1118, %1137, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1139 = torch.aten.mul.Tensor %1133, %arg194 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1140 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1141 = torch.aten.view %1139, %1140 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1142 = torch.aten.t %arg192 : !torch.tensor -> !torch.tensor
    %1143 = torch.aten.mm %1141, %1142 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1144 = torch.aten.t %1141 : !torch.tensor -> !torch.tensor
    %1145 = torch.aten.mm %1144, %arg193 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1146 = torch.aten.t %1145 : !torch.tensor -> !torch.tensor
    %1147 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1148 = torch.aten.sum.dim_IntList %1141, %1147, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1149 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1150 = torch.aten.view %1148, %1149 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1151 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1152 = torch.aten.view %1143, %1151 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1153 = torch.aten.t %1146 : !torch.tensor -> !torch.tensor
    %1154 = torch.aten.gelu_backward %1152, %arg191, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1155 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %1156 = torch.aten.view %1154, %1155 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1157 = torch.aten.t %arg189 : !torch.tensor -> !torch.tensor
    %1158 = torch.aten.mm %1156, %1157 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1159 = torch.aten.t %1156 : !torch.tensor -> !torch.tensor
    %1160 = torch.aten.mm %1159, %arg190 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1161 = torch.aten.t %1160 : !torch.tensor -> !torch.tensor
    %1162 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1163 = torch.aten.sum.dim_IntList %1156, %1162, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1164 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %1165 = torch.aten.view %1163, %1164 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1166 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1167 = torch.aten.view %1158, %1166 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1168 = torch.aten.add.Tensor %1133, %1167, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1169 = torch.aten.t %1161 : !torch.tensor -> !torch.tensor
    %1170 = torch.aten.sub.Tensor %arg188, %result1_31, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1171 = torch.aten.mul.Tensor %1170, %result2_32 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1172 = torch.aten.mul.Tensor %1168, %arg32 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1173 = torch.aten.mul.Tensor %1172, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1174 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1175 = torch.aten.sum.dim_IntList %1172, %1174, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1176 = torch.aten.mul.Tensor %1172, %1171 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1177 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1178 = torch.aten.sum.dim_IntList %1176, %1177, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1179 = torch.aten.mul.Tensor %1171, %1178 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1180 = torch.aten.sub.Tensor %1173, %1175, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1181 = torch.aten.sub.Tensor %1180, %1179, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1182 = torch.aten.div.Tensor %result2_32, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1183 = torch.aten.mul.Tensor %1182, %1181 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1184 = torch.aten.mul.Tensor %1168, %1171 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1185 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1186 = torch.aten.sum.dim_IntList %1184, %1185, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1187 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1188 = torch.aten.sum.dim_IntList %1168, %1187, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1189 = torch.aten.mul.Tensor %1183, %arg187 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1190 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1191 = torch.aten.view %1189, %1190 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1192 = torch.aten.t %arg185 : !torch.tensor -> !torch.tensor
    %1193 = torch.aten.mm %1191, %1192 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1194 = torch.aten.t %1191 : !torch.tensor -> !torch.tensor
    %1195 = torch.aten.mm %1194, %arg186 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1196 = torch.aten.t %1195 : !torch.tensor -> !torch.tensor
    %1197 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1198 = torch.aten.sum.dim_IntList %1191, %1197, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1199 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1200 = torch.aten.view %1198, %1199 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1201 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1202 = torch.aten.view %1193, %1201 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1203 = torch.aten.t %1196 : !torch.tensor -> !torch.tensor
    %1204 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1205 = torch.aten.view %1202, %1204 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1206 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1207 = torch.aten.permute %1205, %1206 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1208 = torch.aten.clone %1207, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1209 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1210 = torch.aten._unsafe_view %1208, %1209 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1211 = torch.aten.transpose.int %arg183, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1212 = torch.aten.bmm %1211, %1210 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1213 = torch.aten.transpose.int %arg184, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1214 = torch.aten.bmm %1210, %1213 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1215 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1216 = torch.aten.view %1212, %1215 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1217 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1218 = torch.aten.view %1214, %1217 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1219 = torch.aten.mul.Tensor %1218, %arg182 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1220 = torch.aten._softmax_backward_data %1219, %arg181, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1221 = torch.aten.div.Tensor %1220, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1222 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1223 = torch.aten.view %1221, %1222 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1224 = torch.aten.transpose.int %arg179, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1225 = torch.aten.bmm %1224, %1223 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1226 = torch.aten.transpose.int %arg180, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1227 = torch.aten.bmm %1223, %1226 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1228 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1229 = torch.aten.view %1225, %1228 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1230 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1231 = torch.aten.view %1227, %1230 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1232 = torch.aten.transpose.int %1229, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1233 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1234 = torch.aten.permute %1231, %1233 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1235 = torch.aten.clone %1234, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1236 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1237 = torch.aten._unsafe_view %1235, %1236 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1238 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1239 = torch.aten.permute %1216, %1238 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1240 = torch.aten.clone %1239, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1241 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1242 = torch.aten._unsafe_view %1240, %1241 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1243 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1244 = torch.aten.view %1242, %1243 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1245 = torch.aten.t %arg177 : !torch.tensor -> !torch.tensor
    %1246 = torch.aten.mm %1244, %1245 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1247 = torch.aten.t %1244 : !torch.tensor -> !torch.tensor
    %1248 = torch.aten.mm %1247, %arg178 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1249 = torch.aten.t %1248 : !torch.tensor -> !torch.tensor
    %1250 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1251 = torch.aten.sum.dim_IntList %1244, %1250, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1252 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1253 = torch.aten.view %1251, %1252 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1254 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1255 = torch.aten.view %1246, %1254 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1256 = torch.aten.add.Tensor %1183, %1255, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1257 = torch.aten.t %1249 : !torch.tensor -> !torch.tensor
    %1258 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1259 = torch.aten.permute %1232, %1258 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1260 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1261 = torch.aten.view %1259, %1260 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1262 = torch.aten.clone %1261, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1263 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1264 = torch.aten._unsafe_view %1262, %1263 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1265 = torch.aten.t %arg175 : !torch.tensor -> !torch.tensor
    %1266 = torch.aten.mm %1264, %1265 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1267 = torch.aten.t %1264 : !torch.tensor -> !torch.tensor
    %1268 = torch.aten.mm %1267, %arg176 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1269 = torch.aten.t %1268 : !torch.tensor -> !torch.tensor
    %1270 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1271 = torch.aten.sum.dim_IntList %1264, %1270, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1272 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1273 = torch.aten.view %1271, %1272 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1274 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1275 = torch.aten.view %1266, %1274 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1276 = torch.aten.add.Tensor %1256, %1275, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1277 = torch.aten.t %1269 : !torch.tensor -> !torch.tensor
    %1278 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1279 = torch.aten.view %1237, %1278 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1280 = torch.aten.t %arg173 : !torch.tensor -> !torch.tensor
    %1281 = torch.aten.mm %1279, %1280 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1282 = torch.aten.t %1279 : !torch.tensor -> !torch.tensor
    %1283 = torch.aten.mm %1282, %arg174 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1284 = torch.aten.t %1283 : !torch.tensor -> !torch.tensor
    %1285 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1286 = torch.aten.sum.dim_IntList %1279, %1285, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1287 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1288 = torch.aten.view %1286, %1287 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1289 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1290 = torch.aten.view %1281, %1289 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1291 = torch.aten.add.Tensor %1276, %1290, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1292 = torch.aten.t %1284 : !torch.tensor -> !torch.tensor
    %1293 = torch.aten.sub.Tensor %arg172, %result1_28, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1294 = torch.aten.mul.Tensor %1293, %result2_29 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1295 = torch.aten.mul.Tensor %1291, %arg30 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1296 = torch.aten.mul.Tensor %1295, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1297 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1298 = torch.aten.sum.dim_IntList %1295, %1297, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1299 = torch.aten.mul.Tensor %1295, %1294 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1300 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1301 = torch.aten.sum.dim_IntList %1299, %1300, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1302 = torch.aten.mul.Tensor %1294, %1301 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1303 = torch.aten.sub.Tensor %1296, %1298, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1304 = torch.aten.sub.Tensor %1303, %1302, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1305 = torch.aten.div.Tensor %result2_29, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1306 = torch.aten.mul.Tensor %1305, %1304 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1307 = torch.aten.mul.Tensor %1291, %1294 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1308 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1309 = torch.aten.sum.dim_IntList %1307, %1308, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1310 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1311 = torch.aten.sum.dim_IntList %1291, %1310, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1312 = torch.aten.mul.Tensor %1306, %arg171 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1313 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1314 = torch.aten.view %1312, %1313 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1315 = torch.aten.t %arg169 : !torch.tensor -> !torch.tensor
    %1316 = torch.aten.mm %1314, %1315 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1317 = torch.aten.t %1314 : !torch.tensor -> !torch.tensor
    %1318 = torch.aten.mm %1317, %arg170 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1319 = torch.aten.t %1318 : !torch.tensor -> !torch.tensor
    %1320 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1321 = torch.aten.sum.dim_IntList %1314, %1320, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1322 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1323 = torch.aten.view %1321, %1322 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1324 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1325 = torch.aten.view %1316, %1324 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1326 = torch.aten.t %1319 : !torch.tensor -> !torch.tensor
    %1327 = torch.aten.gelu_backward %1325, %arg168, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1328 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %1329 = torch.aten.view %1327, %1328 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1330 = torch.aten.t %arg166 : !torch.tensor -> !torch.tensor
    %1331 = torch.aten.mm %1329, %1330 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1332 = torch.aten.t %1329 : !torch.tensor -> !torch.tensor
    %1333 = torch.aten.mm %1332, %arg167 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1334 = torch.aten.t %1333 : !torch.tensor -> !torch.tensor
    %1335 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1336 = torch.aten.sum.dim_IntList %1329, %1335, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1337 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %1338 = torch.aten.view %1336, %1337 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1339 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1340 = torch.aten.view %1331, %1339 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1341 = torch.aten.add.Tensor %1306, %1340, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1342 = torch.aten.t %1334 : !torch.tensor -> !torch.tensor
    %1343 = torch.aten.sub.Tensor %arg165, %result1_25, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1344 = torch.aten.mul.Tensor %1343, %result2_26 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1345 = torch.aten.mul.Tensor %1341, %arg28 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1346 = torch.aten.mul.Tensor %1345, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1347 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1348 = torch.aten.sum.dim_IntList %1345, %1347, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1349 = torch.aten.mul.Tensor %1345, %1344 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1350 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1351 = torch.aten.sum.dim_IntList %1349, %1350, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1352 = torch.aten.mul.Tensor %1344, %1351 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1353 = torch.aten.sub.Tensor %1346, %1348, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1354 = torch.aten.sub.Tensor %1353, %1352, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1355 = torch.aten.div.Tensor %result2_26, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1356 = torch.aten.mul.Tensor %1355, %1354 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1357 = torch.aten.mul.Tensor %1341, %1344 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1358 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1359 = torch.aten.sum.dim_IntList %1357, %1358, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1360 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1361 = torch.aten.sum.dim_IntList %1341, %1360, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1362 = torch.aten.mul.Tensor %1356, %arg164 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1363 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1364 = torch.aten.view %1362, %1363 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1365 = torch.aten.t %arg162 : !torch.tensor -> !torch.tensor
    %1366 = torch.aten.mm %1364, %1365 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1367 = torch.aten.t %1364 : !torch.tensor -> !torch.tensor
    %1368 = torch.aten.mm %1367, %arg163 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1369 = torch.aten.t %1368 : !torch.tensor -> !torch.tensor
    %1370 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1371 = torch.aten.sum.dim_IntList %1364, %1370, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1372 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1373 = torch.aten.view %1371, %1372 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1374 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1375 = torch.aten.view %1366, %1374 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1376 = torch.aten.t %1369 : !torch.tensor -> !torch.tensor
    %1377 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1378 = torch.aten.view %1375, %1377 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1379 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1380 = torch.aten.permute %1378, %1379 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1381 = torch.aten.clone %1380, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1382 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1383 = torch.aten._unsafe_view %1381, %1382 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1384 = torch.aten.transpose.int %arg160, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1385 = torch.aten.bmm %1384, %1383 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1386 = torch.aten.transpose.int %arg161, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1387 = torch.aten.bmm %1383, %1386 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1388 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1389 = torch.aten.view %1385, %1388 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1390 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1391 = torch.aten.view %1387, %1390 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1392 = torch.aten.mul.Tensor %1391, %arg159 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1393 = torch.aten._softmax_backward_data %1392, %arg158, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1394 = torch.aten.div.Tensor %1393, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1395 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1396 = torch.aten.view %1394, %1395 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1397 = torch.aten.transpose.int %arg156, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1398 = torch.aten.bmm %1397, %1396 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1399 = torch.aten.transpose.int %arg157, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1400 = torch.aten.bmm %1396, %1399 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1401 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1402 = torch.aten.view %1398, %1401 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1403 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1404 = torch.aten.view %1400, %1403 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1405 = torch.aten.transpose.int %1402, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1406 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1407 = torch.aten.permute %1404, %1406 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1408 = torch.aten.clone %1407, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1409 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1410 = torch.aten._unsafe_view %1408, %1409 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1411 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1412 = torch.aten.permute %1389, %1411 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1413 = torch.aten.clone %1412, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1414 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1415 = torch.aten._unsafe_view %1413, %1414 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1416 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1417 = torch.aten.view %1415, %1416 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1418 = torch.aten.t %arg154 : !torch.tensor -> !torch.tensor
    %1419 = torch.aten.mm %1417, %1418 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1420 = torch.aten.t %1417 : !torch.tensor -> !torch.tensor
    %1421 = torch.aten.mm %1420, %arg155 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1422 = torch.aten.t %1421 : !torch.tensor -> !torch.tensor
    %1423 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1424 = torch.aten.sum.dim_IntList %1417, %1423, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1425 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1426 = torch.aten.view %1424, %1425 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1427 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1428 = torch.aten.view %1419, %1427 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1429 = torch.aten.add.Tensor %1356, %1428, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1430 = torch.aten.t %1422 : !torch.tensor -> !torch.tensor
    %1431 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1432 = torch.aten.permute %1405, %1431 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1433 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1434 = torch.aten.view %1432, %1433 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1435 = torch.aten.clone %1434, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1436 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1437 = torch.aten._unsafe_view %1435, %1436 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1438 = torch.aten.t %arg152 : !torch.tensor -> !torch.tensor
    %1439 = torch.aten.mm %1437, %1438 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1440 = torch.aten.t %1437 : !torch.tensor -> !torch.tensor
    %1441 = torch.aten.mm %1440, %arg153 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1442 = torch.aten.t %1441 : !torch.tensor -> !torch.tensor
    %1443 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1444 = torch.aten.sum.dim_IntList %1437, %1443, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1445 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1446 = torch.aten.view %1444, %1445 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1447 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1448 = torch.aten.view %1439, %1447 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1449 = torch.aten.add.Tensor %1429, %1448, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1450 = torch.aten.t %1442 : !torch.tensor -> !torch.tensor
    %1451 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1452 = torch.aten.view %1410, %1451 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1453 = torch.aten.t %arg150 : !torch.tensor -> !torch.tensor
    %1454 = torch.aten.mm %1452, %1453 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1455 = torch.aten.t %1452 : !torch.tensor -> !torch.tensor
    %1456 = torch.aten.mm %1455, %arg151 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1457 = torch.aten.t %1456 : !torch.tensor -> !torch.tensor
    %1458 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1459 = torch.aten.sum.dim_IntList %1452, %1458, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1460 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1461 = torch.aten.view %1459, %1460 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1462 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1463 = torch.aten.view %1454, %1462 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1464 = torch.aten.add.Tensor %1449, %1463, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1465 = torch.aten.t %1457 : !torch.tensor -> !torch.tensor
    %1466 = torch.aten.sub.Tensor %arg149, %result1_22, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1467 = torch.aten.mul.Tensor %1466, %result2_23 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1468 = torch.aten.mul.Tensor %1464, %arg26 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1469 = torch.aten.mul.Tensor %1468, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1470 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1471 = torch.aten.sum.dim_IntList %1468, %1470, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1472 = torch.aten.mul.Tensor %1468, %1467 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1473 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1474 = torch.aten.sum.dim_IntList %1472, %1473, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1475 = torch.aten.mul.Tensor %1467, %1474 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1476 = torch.aten.sub.Tensor %1469, %1471, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1477 = torch.aten.sub.Tensor %1476, %1475, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1478 = torch.aten.div.Tensor %result2_23, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1479 = torch.aten.mul.Tensor %1478, %1477 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1480 = torch.aten.mul.Tensor %1464, %1467 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1481 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1482 = torch.aten.sum.dim_IntList %1480, %1481, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1483 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1484 = torch.aten.sum.dim_IntList %1464, %1483, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1485 = torch.aten.mul.Tensor %1479, %arg148 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1486 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1487 = torch.aten.view %1485, %1486 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1488 = torch.aten.t %arg146 : !torch.tensor -> !torch.tensor
    %1489 = torch.aten.mm %1487, %1488 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1490 = torch.aten.t %1487 : !torch.tensor -> !torch.tensor
    %1491 = torch.aten.mm %1490, %arg147 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1492 = torch.aten.t %1491 : !torch.tensor -> !torch.tensor
    %1493 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1494 = torch.aten.sum.dim_IntList %1487, %1493, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1495 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1496 = torch.aten.view %1494, %1495 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1497 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1498 = torch.aten.view %1489, %1497 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1499 = torch.aten.t %1492 : !torch.tensor -> !torch.tensor
    %1500 = torch.aten.gelu_backward %1498, %arg145, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1501 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %1502 = torch.aten.view %1500, %1501 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1503 = torch.aten.t %arg143 : !torch.tensor -> !torch.tensor
    %1504 = torch.aten.mm %1502, %1503 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1505 = torch.aten.t %1502 : !torch.tensor -> !torch.tensor
    %1506 = torch.aten.mm %1505, %arg144 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1507 = torch.aten.t %1506 : !torch.tensor -> !torch.tensor
    %1508 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1509 = torch.aten.sum.dim_IntList %1502, %1508, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1510 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %1511 = torch.aten.view %1509, %1510 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1512 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1513 = torch.aten.view %1504, %1512 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1514 = torch.aten.add.Tensor %1479, %1513, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1515 = torch.aten.t %1507 : !torch.tensor -> !torch.tensor
    %1516 = torch.aten.sub.Tensor %arg142, %result1_19, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1517 = torch.aten.mul.Tensor %1516, %result2_20 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1518 = torch.aten.mul.Tensor %1514, %arg24 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1519 = torch.aten.mul.Tensor %1518, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1520 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1521 = torch.aten.sum.dim_IntList %1518, %1520, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1522 = torch.aten.mul.Tensor %1518, %1517 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1523 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1524 = torch.aten.sum.dim_IntList %1522, %1523, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1525 = torch.aten.mul.Tensor %1517, %1524 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1526 = torch.aten.sub.Tensor %1519, %1521, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1527 = torch.aten.sub.Tensor %1526, %1525, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1528 = torch.aten.div.Tensor %result2_20, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1529 = torch.aten.mul.Tensor %1528, %1527 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1530 = torch.aten.mul.Tensor %1514, %1517 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1531 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1532 = torch.aten.sum.dim_IntList %1530, %1531, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1533 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1534 = torch.aten.sum.dim_IntList %1514, %1533, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1535 = torch.aten.mul.Tensor %1529, %arg141 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1536 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1537 = torch.aten.view %1535, %1536 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1538 = torch.aten.t %arg139 : !torch.tensor -> !torch.tensor
    %1539 = torch.aten.mm %1537, %1538 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1540 = torch.aten.t %1537 : !torch.tensor -> !torch.tensor
    %1541 = torch.aten.mm %1540, %arg140 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1542 = torch.aten.t %1541 : !torch.tensor -> !torch.tensor
    %1543 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1544 = torch.aten.sum.dim_IntList %1537, %1543, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1545 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1546 = torch.aten.view %1544, %1545 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1547 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1548 = torch.aten.view %1539, %1547 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1549 = torch.aten.t %1542 : !torch.tensor -> !torch.tensor
    %1550 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1551 = torch.aten.view %1548, %1550 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1552 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1553 = torch.aten.permute %1551, %1552 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1554 = torch.aten.clone %1553, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1555 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1556 = torch.aten._unsafe_view %1554, %1555 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1557 = torch.aten.transpose.int %arg137, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1558 = torch.aten.bmm %1557, %1556 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1559 = torch.aten.transpose.int %arg138, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1560 = torch.aten.bmm %1556, %1559 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1561 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1562 = torch.aten.view %1558, %1561 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1563 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1564 = torch.aten.view %1560, %1563 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1565 = torch.aten.mul.Tensor %1564, %arg136 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1566 = torch.aten._softmax_backward_data %1565, %arg135, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1567 = torch.aten.div.Tensor %1566, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1568 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1569 = torch.aten.view %1567, %1568 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1570 = torch.aten.transpose.int %arg133, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1571 = torch.aten.bmm %1570, %1569 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1572 = torch.aten.transpose.int %arg134, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1573 = torch.aten.bmm %1569, %1572 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1574 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1575 = torch.aten.view %1571, %1574 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1576 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1577 = torch.aten.view %1573, %1576 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1578 = torch.aten.transpose.int %1575, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1579 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1580 = torch.aten.permute %1577, %1579 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1581 = torch.aten.clone %1580, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1582 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1583 = torch.aten._unsafe_view %1581, %1582 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1584 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1585 = torch.aten.permute %1562, %1584 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1586 = torch.aten.clone %1585, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1587 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1588 = torch.aten._unsafe_view %1586, %1587 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1589 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1590 = torch.aten.view %1588, %1589 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1591 = torch.aten.t %arg131 : !torch.tensor -> !torch.tensor
    %1592 = torch.aten.mm %1590, %1591 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1593 = torch.aten.t %1590 : !torch.tensor -> !torch.tensor
    %1594 = torch.aten.mm %1593, %arg132 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1595 = torch.aten.t %1594 : !torch.tensor -> !torch.tensor
    %1596 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1597 = torch.aten.sum.dim_IntList %1590, %1596, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1598 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1599 = torch.aten.view %1597, %1598 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1600 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1601 = torch.aten.view %1592, %1600 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1602 = torch.aten.add.Tensor %1529, %1601, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1603 = torch.aten.t %1595 : !torch.tensor -> !torch.tensor
    %1604 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1605 = torch.aten.permute %1578, %1604 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1606 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1607 = torch.aten.view %1605, %1606 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1608 = torch.aten.clone %1607, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1609 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1610 = torch.aten._unsafe_view %1608, %1609 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1611 = torch.aten.t %arg129 : !torch.tensor -> !torch.tensor
    %1612 = torch.aten.mm %1610, %1611 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1613 = torch.aten.t %1610 : !torch.tensor -> !torch.tensor
    %1614 = torch.aten.mm %1613, %arg130 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1615 = torch.aten.t %1614 : !torch.tensor -> !torch.tensor
    %1616 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1617 = torch.aten.sum.dim_IntList %1610, %1616, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1618 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1619 = torch.aten.view %1617, %1618 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1620 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1621 = torch.aten.view %1612, %1620 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1622 = torch.aten.add.Tensor %1602, %1621, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1623 = torch.aten.t %1615 : !torch.tensor -> !torch.tensor
    %1624 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1625 = torch.aten.view %1583, %1624 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1626 = torch.aten.t %arg127 : !torch.tensor -> !torch.tensor
    %1627 = torch.aten.mm %1625, %1626 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1628 = torch.aten.t %1625 : !torch.tensor -> !torch.tensor
    %1629 = torch.aten.mm %1628, %arg128 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1630 = torch.aten.t %1629 : !torch.tensor -> !torch.tensor
    %1631 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1632 = torch.aten.sum.dim_IntList %1625, %1631, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1633 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1634 = torch.aten.view %1632, %1633 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1635 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1636 = torch.aten.view %1627, %1635 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1637 = torch.aten.add.Tensor %1622, %1636, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1638 = torch.aten.t %1630 : !torch.tensor -> !torch.tensor
    %1639 = torch.aten.sub.Tensor %arg126, %result1_16, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1640 = torch.aten.mul.Tensor %1639, %result2_17 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1641 = torch.aten.mul.Tensor %1637, %arg22 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1642 = torch.aten.mul.Tensor %1641, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1643 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1644 = torch.aten.sum.dim_IntList %1641, %1643, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1645 = torch.aten.mul.Tensor %1641, %1640 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1646 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1647 = torch.aten.sum.dim_IntList %1645, %1646, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1648 = torch.aten.mul.Tensor %1640, %1647 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1649 = torch.aten.sub.Tensor %1642, %1644, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1650 = torch.aten.sub.Tensor %1649, %1648, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1651 = torch.aten.div.Tensor %result2_17, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1652 = torch.aten.mul.Tensor %1651, %1650 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1653 = torch.aten.mul.Tensor %1637, %1640 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1654 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1655 = torch.aten.sum.dim_IntList %1653, %1654, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1656 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1657 = torch.aten.sum.dim_IntList %1637, %1656, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1658 = torch.aten.mul.Tensor %1652, %arg125 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1659 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1660 = torch.aten.view %1658, %1659 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1661 = torch.aten.t %arg123 : !torch.tensor -> !torch.tensor
    %1662 = torch.aten.mm %1660, %1661 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1663 = torch.aten.t %1660 : !torch.tensor -> !torch.tensor
    %1664 = torch.aten.mm %1663, %arg124 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1665 = torch.aten.t %1664 : !torch.tensor -> !torch.tensor
    %1666 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1667 = torch.aten.sum.dim_IntList %1660, %1666, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1668 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1669 = torch.aten.view %1667, %1668 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1670 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1671 = torch.aten.view %1662, %1670 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1672 = torch.aten.t %1665 : !torch.tensor -> !torch.tensor
    %1673 = torch.aten.gelu_backward %1671, %arg122, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1674 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %1675 = torch.aten.view %1673, %1674 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1676 = torch.aten.t %arg120 : !torch.tensor -> !torch.tensor
    %1677 = torch.aten.mm %1675, %1676 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1678 = torch.aten.t %1675 : !torch.tensor -> !torch.tensor
    %1679 = torch.aten.mm %1678, %arg121 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1680 = torch.aten.t %1679 : !torch.tensor -> !torch.tensor
    %1681 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1682 = torch.aten.sum.dim_IntList %1675, %1681, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1683 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %1684 = torch.aten.view %1682, %1683 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1685 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1686 = torch.aten.view %1677, %1685 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1687 = torch.aten.add.Tensor %1652, %1686, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1688 = torch.aten.t %1680 : !torch.tensor -> !torch.tensor
    %1689 = torch.aten.sub.Tensor %arg119, %result1_13, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1690 = torch.aten.mul.Tensor %1689, %result2_14 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1691 = torch.aten.mul.Tensor %1687, %arg20 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1692 = torch.aten.mul.Tensor %1691, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1693 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1694 = torch.aten.sum.dim_IntList %1691, %1693, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1695 = torch.aten.mul.Tensor %1691, %1690 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1696 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1697 = torch.aten.sum.dim_IntList %1695, %1696, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1698 = torch.aten.mul.Tensor %1690, %1697 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1699 = torch.aten.sub.Tensor %1692, %1694, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1700 = torch.aten.sub.Tensor %1699, %1698, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1701 = torch.aten.div.Tensor %result2_14, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1702 = torch.aten.mul.Tensor %1701, %1700 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1703 = torch.aten.mul.Tensor %1687, %1690 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1704 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1705 = torch.aten.sum.dim_IntList %1703, %1704, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1706 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1707 = torch.aten.sum.dim_IntList %1687, %1706, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1708 = torch.aten.mul.Tensor %1702, %arg118 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1709 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1710 = torch.aten.view %1708, %1709 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1711 = torch.aten.t %arg116 : !torch.tensor -> !torch.tensor
    %1712 = torch.aten.mm %1710, %1711 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1713 = torch.aten.t %1710 : !torch.tensor -> !torch.tensor
    %1714 = torch.aten.mm %1713, %arg117 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1715 = torch.aten.t %1714 : !torch.tensor -> !torch.tensor
    %1716 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1717 = torch.aten.sum.dim_IntList %1710, %1716, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1718 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1719 = torch.aten.view %1717, %1718 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1720 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1721 = torch.aten.view %1712, %1720 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1722 = torch.aten.t %1715 : !torch.tensor -> !torch.tensor
    %1723 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1724 = torch.aten.view %1721, %1723 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1725 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1726 = torch.aten.permute %1724, %1725 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1727 = torch.aten.clone %1726, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1728 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1729 = torch.aten._unsafe_view %1727, %1728 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1730 = torch.aten.transpose.int %arg114, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1731 = torch.aten.bmm %1730, %1729 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1732 = torch.aten.transpose.int %arg115, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1733 = torch.aten.bmm %1729, %1732 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1734 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1735 = torch.aten.view %1731, %1734 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1736 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1737 = torch.aten.view %1733, %1736 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1738 = torch.aten.mul.Tensor %1737, %arg113 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1739 = torch.aten._softmax_backward_data %1738, %arg112, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1740 = torch.aten.div.Tensor %1739, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1741 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1742 = torch.aten.view %1740, %1741 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1743 = torch.aten.transpose.int %arg110, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1744 = torch.aten.bmm %1743, %1742 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1745 = torch.aten.transpose.int %arg111, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1746 = torch.aten.bmm %1742, %1745 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1747 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1748 = torch.aten.view %1744, %1747 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1749 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1750 = torch.aten.view %1746, %1749 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1751 = torch.aten.transpose.int %1748, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1752 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1753 = torch.aten.permute %1750, %1752 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1754 = torch.aten.clone %1753, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1755 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1756 = torch.aten._unsafe_view %1754, %1755 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1757 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1758 = torch.aten.permute %1735, %1757 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1759 = torch.aten.clone %1758, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1760 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1761 = torch.aten._unsafe_view %1759, %1760 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1762 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1763 = torch.aten.view %1761, %1762 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1764 = torch.aten.t %arg108 : !torch.tensor -> !torch.tensor
    %1765 = torch.aten.mm %1763, %1764 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1766 = torch.aten.t %1763 : !torch.tensor -> !torch.tensor
    %1767 = torch.aten.mm %1766, %arg109 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1768 = torch.aten.t %1767 : !torch.tensor -> !torch.tensor
    %1769 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1770 = torch.aten.sum.dim_IntList %1763, %1769, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1771 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1772 = torch.aten.view %1770, %1771 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1773 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1774 = torch.aten.view %1765, %1773 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1775 = torch.aten.add.Tensor %1702, %1774, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1776 = torch.aten.t %1768 : !torch.tensor -> !torch.tensor
    %1777 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1778 = torch.aten.permute %1751, %1777 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1779 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1780 = torch.aten.view %1778, %1779 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1781 = torch.aten.clone %1780, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1782 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1783 = torch.aten._unsafe_view %1781, %1782 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1784 = torch.aten.t %arg106 : !torch.tensor -> !torch.tensor
    %1785 = torch.aten.mm %1783, %1784 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1786 = torch.aten.t %1783 : !torch.tensor -> !torch.tensor
    %1787 = torch.aten.mm %1786, %arg107 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1788 = torch.aten.t %1787 : !torch.tensor -> !torch.tensor
    %1789 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1790 = torch.aten.sum.dim_IntList %1783, %1789, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1791 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1792 = torch.aten.view %1790, %1791 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1793 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1794 = torch.aten.view %1785, %1793 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1795 = torch.aten.add.Tensor %1775, %1794, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1796 = torch.aten.t %1788 : !torch.tensor -> !torch.tensor
    %1797 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1798 = torch.aten.view %1756, %1797 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1799 = torch.aten.t %arg104 : !torch.tensor -> !torch.tensor
    %1800 = torch.aten.mm %1798, %1799 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1801 = torch.aten.t %1798 : !torch.tensor -> !torch.tensor
    %1802 = torch.aten.mm %1801, %arg105 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1803 = torch.aten.t %1802 : !torch.tensor -> !torch.tensor
    %1804 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1805 = torch.aten.sum.dim_IntList %1798, %1804, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1806 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1807 = torch.aten.view %1805, %1806 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1808 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1809 = torch.aten.view %1800, %1808 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1810 = torch.aten.add.Tensor %1795, %1809, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1811 = torch.aten.t %1803 : !torch.tensor -> !torch.tensor
    %1812 = torch.aten.sub.Tensor %arg103, %result1_10, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1813 = torch.aten.mul.Tensor %1812, %result2_11 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1814 = torch.aten.mul.Tensor %1810, %arg10 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1815 = torch.aten.mul.Tensor %1814, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1816 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1817 = torch.aten.sum.dim_IntList %1814, %1816, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1818 = torch.aten.mul.Tensor %1814, %1813 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1819 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1820 = torch.aten.sum.dim_IntList %1818, %1819, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1821 = torch.aten.mul.Tensor %1813, %1820 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1822 = torch.aten.sub.Tensor %1815, %1817, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1823 = torch.aten.sub.Tensor %1822, %1821, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1824 = torch.aten.div.Tensor %result2_11, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1825 = torch.aten.mul.Tensor %1824, %1823 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1826 = torch.aten.mul.Tensor %1810, %1813 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1827 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1828 = torch.aten.sum.dim_IntList %1826, %1827, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1829 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1830 = torch.aten.sum.dim_IntList %1810, %1829, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1831 = torch.aten.mul.Tensor %1825, %arg102 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1832 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1833 = torch.aten.view %1831, %1832 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1834 = torch.aten.t %arg100 : !torch.tensor -> !torch.tensor
    %1835 = torch.aten.mm %1833, %1834 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1836 = torch.aten.t %1833 : !torch.tensor -> !torch.tensor
    %1837 = torch.aten.mm %1836, %arg101 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1838 = torch.aten.t %1837 : !torch.tensor -> !torch.tensor
    %1839 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1840 = torch.aten.sum.dim_IntList %1833, %1839, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1841 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1842 = torch.aten.view %1840, %1841 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1843 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1844 = torch.aten.view %1835, %1843 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1845 = torch.aten.t %1838 : !torch.tensor -> !torch.tensor
    %1846 = torch.aten.gelu_backward %1844, %arg99, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %1847 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %1848 = torch.aten.view %1846, %1847 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1849 = torch.aten.t %arg97 : !torch.tensor -> !torch.tensor
    %1850 = torch.aten.mm %1848, %1849 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1851 = torch.aten.t %1848 : !torch.tensor -> !torch.tensor
    %1852 = torch.aten.mm %1851, %arg98 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1853 = torch.aten.t %1852 : !torch.tensor -> !torch.tensor
    %1854 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1855 = torch.aten.sum.dim_IntList %1848, %1854, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1856 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %1857 = torch.aten.view %1855, %1856 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1858 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1859 = torch.aten.view %1850, %1858 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1860 = torch.aten.add.Tensor %1825, %1859, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1861 = torch.aten.t %1853 : !torch.tensor -> !torch.tensor
    %1862 = torch.aten.sub.Tensor %arg96, %result1_7, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1863 = torch.aten.mul.Tensor %1862, %result2_8 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1864 = torch.aten.mul.Tensor %1860, %arg8 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1865 = torch.aten.mul.Tensor %1864, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1866 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1867 = torch.aten.sum.dim_IntList %1864, %1866, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1868 = torch.aten.mul.Tensor %1864, %1863 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1869 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1870 = torch.aten.sum.dim_IntList %1868, %1869, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1871 = torch.aten.mul.Tensor %1863, %1870 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1872 = torch.aten.sub.Tensor %1865, %1867, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1873 = torch.aten.sub.Tensor %1872, %1871, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1874 = torch.aten.div.Tensor %result2_8, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1875 = torch.aten.mul.Tensor %1874, %1873 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1876 = torch.aten.mul.Tensor %1860, %1863 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1877 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1878 = torch.aten.sum.dim_IntList %1876, %1877, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1879 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %1880 = torch.aten.sum.dim_IntList %1860, %1879, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1881 = torch.aten.mul.Tensor %1875, %arg95 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1882 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1883 = torch.aten.view %1881, %1882 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1884 = torch.aten.t %arg93 : !torch.tensor -> !torch.tensor
    %1885 = torch.aten.mm %1883, %1884 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1886 = torch.aten.t %1883 : !torch.tensor -> !torch.tensor
    %1887 = torch.aten.mm %1886, %arg94 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1888 = torch.aten.t %1887 : !torch.tensor -> !torch.tensor
    %1889 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1890 = torch.aten.sum.dim_IntList %1883, %1889, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1891 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1892 = torch.aten.view %1890, %1891 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1893 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1894 = torch.aten.view %1885, %1893 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1895 = torch.aten.t %1888 : !torch.tensor -> !torch.tensor
    %1896 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1897 = torch.aten.view %1894, %1896 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1898 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1899 = torch.aten.permute %1897, %1898 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1900 = torch.aten.clone %1899, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1901 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1902 = torch.aten._unsafe_view %1900, %1901 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1903 = torch.aten.transpose.int %arg91, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1904 = torch.aten.bmm %1903, %1902 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1905 = torch.aten.transpose.int %arg92, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1906 = torch.aten.bmm %1902, %1905 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1907 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1908 = torch.aten.view %1904, %1907 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1909 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1910 = torch.aten.view %1906, %1909 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1911 = torch.aten.mul.Tensor %1910, %arg90 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1912 = torch.aten._softmax_backward_data %1911, %arg89, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1913 = torch.aten.div.Tensor %1912, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %1914 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1915 = torch.aten.view %1913, %1914 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1916 = torch.aten.transpose.int %arg87, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1917 = torch.aten.bmm %1916, %1915 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1918 = torch.aten.transpose.int %arg88, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1919 = torch.aten.bmm %1915, %1918 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1920 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1921 = torch.aten.view %1917, %1920 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1922 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1923 = torch.aten.view %1919, %1922 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1924 = torch.aten.transpose.int %1921, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %1925 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1926 = torch.aten.permute %1923, %1925 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1927 = torch.aten.clone %1926, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1928 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1929 = torch.aten._unsafe_view %1927, %1928 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1930 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1931 = torch.aten.permute %1908, %1930 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1932 = torch.aten.clone %1931, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1933 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1934 = torch.aten._unsafe_view %1932, %1933 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1935 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1936 = torch.aten.view %1934, %1935 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1937 = torch.aten.t %arg85 : !torch.tensor -> !torch.tensor
    %1938 = torch.aten.mm %1936, %1937 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1939 = torch.aten.t %1936 : !torch.tensor -> !torch.tensor
    %1940 = torch.aten.mm %1939, %arg86 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1941 = torch.aten.t %1940 : !torch.tensor -> !torch.tensor
    %1942 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1943 = torch.aten.sum.dim_IntList %1936, %1942, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1944 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1945 = torch.aten.view %1943, %1944 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1946 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1947 = torch.aten.view %1938, %1946 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1948 = torch.aten.add.Tensor %1875, %1947, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1949 = torch.aten.t %1941 : !torch.tensor -> !torch.tensor
    %1950 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1951 = torch.aten.permute %1924, %1950 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1952 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1953 = torch.aten.view %1951, %1952 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1954 = torch.aten.clone %1953, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %1955 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1956 = torch.aten._unsafe_view %1954, %1955 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1957 = torch.aten.t %arg83 : !torch.tensor -> !torch.tensor
    %1958 = torch.aten.mm %1956, %1957 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1959 = torch.aten.t %1956 : !torch.tensor -> !torch.tensor
    %1960 = torch.aten.mm %1959, %arg84 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1961 = torch.aten.t %1960 : !torch.tensor -> !torch.tensor
    %1962 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1963 = torch.aten.sum.dim_IntList %1956, %1962, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1964 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1965 = torch.aten.view %1963, %1964 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1966 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1967 = torch.aten.view %1958, %1966 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1968 = torch.aten.add.Tensor %1948, %1967, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1969 = torch.aten.t %1961 : !torch.tensor -> !torch.tensor
    %1970 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %1971 = torch.aten.view %1929, %1970 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1972 = torch.aten.t %arg81 : !torch.tensor -> !torch.tensor
    %1973 = torch.aten.mm %1971, %1972 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1974 = torch.aten.t %1971 : !torch.tensor -> !torch.tensor
    %1975 = torch.aten.mm %1974, %arg82 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1976 = torch.aten.t %1975 : !torch.tensor -> !torch.tensor
    %1977 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %1978 = torch.aten.sum.dim_IntList %1971, %1977, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1979 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %1980 = torch.aten.view %1978, %1979 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1981 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %1982 = torch.aten.view %1973, %1981 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %1983 = torch.aten.add.Tensor %1968, %1982, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1984 = torch.aten.t %1976 : !torch.tensor -> !torch.tensor
    %1985 = torch.aten.sub.Tensor %arg80, %result1_4, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1986 = torch.aten.mul.Tensor %1985, %result2_5 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1987 = torch.aten.mul.Tensor %1983, %arg6 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1988 = torch.aten.mul.Tensor %1987, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1989 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1990 = torch.aten.sum.dim_IntList %1987, %1989, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1991 = torch.aten.mul.Tensor %1987, %1986 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1992 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %1993 = torch.aten.sum.dim_IntList %1991, %1992, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %1994 = torch.aten.mul.Tensor %1986, %1993 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1995 = torch.aten.sub.Tensor %1988, %1990, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1996 = torch.aten.sub.Tensor %1995, %1994, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %1997 = torch.aten.div.Tensor %result2_5, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %1998 = torch.aten.mul.Tensor %1997, %1996 : !torch.tensor, !torch.tensor -> !torch.tensor
    %1999 = torch.aten.mul.Tensor %1983, %1986 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2000 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2001 = torch.aten.sum.dim_IntList %1999, %2000, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2002 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2003 = torch.aten.sum.dim_IntList %1983, %2002, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2004 = torch.aten.mul.Tensor %1998, %arg79 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2005 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2006 = torch.aten.view %2004, %2005 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2007 = torch.aten.t %arg77 : !torch.tensor -> !torch.tensor
    %2008 = torch.aten.mm %2006, %2007 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2009 = torch.aten.t %2006 : !torch.tensor -> !torch.tensor
    %2010 = torch.aten.mm %2009, %arg78 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2011 = torch.aten.t %2010 : !torch.tensor -> !torch.tensor
    %2012 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2013 = torch.aten.sum.dim_IntList %2006, %2012, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2014 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %2015 = torch.aten.view %2013, %2014 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2016 = torch.prim.ListConstruct %int4, %int512, %int3072 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2017 = torch.aten.view %2008, %2016 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2018 = torch.aten.t %2011 : !torch.tensor -> !torch.tensor
    %2019 = torch.aten.gelu_backward %2017, %arg76, %str : !torch.tensor, !torch.tensor, !torch.str -> !torch.tensor
    %2020 = torch.prim.ListConstruct %int2048, %int3072 : (!torch.int, !torch.int) -> !torch.list<int>
    %2021 = torch.aten.view %2019, %2020 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2022 = torch.aten.t %arg74 : !torch.tensor -> !torch.tensor
    %2023 = torch.aten.mm %2021, %2022 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2024 = torch.aten.t %2021 : !torch.tensor -> !torch.tensor
    %2025 = torch.aten.mm %2024, %arg75 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2026 = torch.aten.t %2025 : !torch.tensor -> !torch.tensor
    %2027 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2028 = torch.aten.sum.dim_IntList %2021, %2027, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2029 = torch.prim.ListConstruct %int3072 : (!torch.int) -> !torch.list<int>
    %2030 = torch.aten.view %2028, %2029 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2031 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2032 = torch.aten.view %2023, %2031 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2033 = torch.aten.add.Tensor %1998, %2032, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2034 = torch.aten.t %2026 : !torch.tensor -> !torch.tensor
    %2035 = torch.aten.sub.Tensor %arg73, %result1_1, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2036 = torch.aten.mul.Tensor %2035, %result2_2 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2037 = torch.aten.mul.Tensor %2033, %arg4 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2038 = torch.aten.mul.Tensor %2037, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %2039 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2040 = torch.aten.sum.dim_IntList %2037, %2039, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2041 = torch.aten.mul.Tensor %2037, %2036 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2042 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2043 = torch.aten.sum.dim_IntList %2041, %2042, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2044 = torch.aten.mul.Tensor %2036, %2043 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2045 = torch.aten.sub.Tensor %2038, %2040, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2046 = torch.aten.sub.Tensor %2045, %2044, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2047 = torch.aten.div.Tensor %result2_2, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %2048 = torch.aten.mul.Tensor %2047, %2046 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2049 = torch.aten.mul.Tensor %2033, %2036 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2050 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2051 = torch.aten.sum.dim_IntList %2049, %2050, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2052 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2053 = torch.aten.sum.dim_IntList %2033, %2052, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2054 = torch.aten.mul.Tensor %2048, %arg72 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2055 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2056 = torch.aten.view %2054, %2055 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2057 = torch.aten.t %arg70 : !torch.tensor -> !torch.tensor
    %2058 = torch.aten.mm %2056, %2057 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2059 = torch.aten.t %2056 : !torch.tensor -> !torch.tensor
    %2060 = torch.aten.mm %2059, %arg71 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2061 = torch.aten.t %2060 : !torch.tensor -> !torch.tensor
    %2062 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2063 = torch.aten.sum.dim_IntList %2056, %2062, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2064 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %2065 = torch.aten.view %2063, %2064 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2066 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2067 = torch.aten.view %2058, %2066 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2068 = torch.aten.t %2061 : !torch.tensor -> !torch.tensor
    %2069 = torch.prim.ListConstruct %int4, %int512, %int12, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2070 = torch.aten.view %2067, %2069 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2071 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2072 = torch.aten.permute %2070, %2071 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2073 = torch.aten.clone %2072, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2074 = torch.prim.ListConstruct %int48, %int512, %int64 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2075 = torch.aten._unsafe_view %2073, %2074 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2076 = torch.aten.transpose.int %arg68, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2077 = torch.aten.bmm %2076, %2075 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2078 = torch.aten.transpose.int %arg69, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2079 = torch.aten.bmm %2075, %2078 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2080 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2081 = torch.aten.view %2077, %2080 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2082 = torch.prim.ListConstruct %int4, %int12, %int512, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2083 = torch.aten.view %2079, %2082 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2084 = torch.aten.mul.Tensor %2083, %arg67 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2085 = torch.aten._softmax_backward_data %2084, %arg66, %int-1, %int6 : !torch.tensor, !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2086 = torch.aten.div.Tensor %2085, %1 : !torch.tensor, !torch.tensor<[],f64> -> !torch.tensor
    %2087 = torch.prim.ListConstruct %int48, %int512, %int512 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2088 = torch.aten.view %2086, %2087 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2089 = torch.aten.transpose.int %arg64, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2090 = torch.aten.bmm %2089, %2088 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2091 = torch.aten.transpose.int %arg65, %int1, %int2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2092 = torch.aten.bmm %2088, %2091 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2093 = torch.prim.ListConstruct %int4, %int12, %int64, %int512 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2094 = torch.aten.view %2090, %2093 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2095 = torch.prim.ListConstruct %int4, %int12, %int512, %int64 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2096 = torch.aten.view %2092, %2095 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2097 = torch.aten.transpose.int %2094, %int-1, %int-2 : !torch.tensor, !torch.int, !torch.int -> !torch.tensor
    %2098 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2099 = torch.aten.permute %2096, %2098 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2100 = torch.aten.clone %2099, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2101 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2102 = torch.aten._unsafe_view %2100, %2101 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2103 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2104 = torch.aten.permute %2081, %2103 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2105 = torch.aten.clone %2104, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2106 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2107 = torch.aten._unsafe_view %2105, %2106 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2108 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2109 = torch.aten.view %2107, %2108 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2110 = torch.aten.t %arg62 : !torch.tensor -> !torch.tensor
    %2111 = torch.aten.mm %2109, %2110 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2112 = torch.aten.t %2109 : !torch.tensor -> !torch.tensor
    %2113 = torch.aten.mm %2112, %arg63 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2114 = torch.aten.t %2113 : !torch.tensor -> !torch.tensor
    %2115 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2116 = torch.aten.sum.dim_IntList %2109, %2115, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2117 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %2118 = torch.aten.view %2116, %2117 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2119 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2120 = torch.aten.view %2111, %2119 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2121 = torch.aten.add.Tensor %2048, %2120, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2122 = torch.aten.t %2114 : !torch.tensor -> !torch.tensor
    %2123 = torch.prim.ListConstruct %int0, %int2, %int1, %int3 : (!torch.int, !torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2124 = torch.aten.permute %2097, %2123 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2125 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2126 = torch.aten.view %2124, %2125 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2127 = torch.aten.clone %2126, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2128 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2129 = torch.aten._unsafe_view %2127, %2128 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2130 = torch.aten.t %arg60 : !torch.tensor -> !torch.tensor
    %2131 = torch.aten.mm %2129, %2130 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2132 = torch.aten.t %2129 : !torch.tensor -> !torch.tensor
    %2133 = torch.aten.mm %2132, %arg61 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2134 = torch.aten.t %2133 : !torch.tensor -> !torch.tensor
    %2135 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2136 = torch.aten.sum.dim_IntList %2129, %2135, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2137 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %2138 = torch.aten.view %2136, %2137 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2139 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2140 = torch.aten.view %2131, %2139 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2141 = torch.aten.add.Tensor %2121, %2140, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2142 = torch.aten.t %2134 : !torch.tensor -> !torch.tensor
    %2143 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2144 = torch.aten.view %2102, %2143 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2145 = torch.aten.t %arg58 : !torch.tensor -> !torch.tensor
    %2146 = torch.aten.mm %2144, %2145 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2147 = torch.aten.t %2144 : !torch.tensor -> !torch.tensor
    %2148 = torch.aten.mm %2147, %arg59 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2149 = torch.aten.t %2148 : !torch.tensor -> !torch.tensor
    %2150 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2151 = torch.aten.sum.dim_IntList %2144, %2150, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2152 = torch.prim.ListConstruct %int768 : (!torch.int) -> !torch.list<int>
    %2153 = torch.aten.view %2151, %2152 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2154 = torch.prim.ListConstruct %int4, %int512, %int768 : (!torch.int, !torch.int, !torch.int) -> !torch.list<int>
    %2155 = torch.aten.view %2146, %2154 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2156 = torch.aten.add.Tensor %2141, %2155, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2157 = torch.aten.t %2149 : !torch.tensor -> !torch.tensor
    %2158 = torch.aten.mul.Tensor %2156, %arg57 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2159 = torch.aten.sub.Tensor %arg56, %result1, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2160 = torch.aten.mul.Tensor %2159, %result2 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2161 = torch.aten.mul.Tensor %2158, %arg2 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2162 = torch.aten.mul.Tensor %2161, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %2163 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2164 = torch.aten.sum.dim_IntList %2161, %2163, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2165 = torch.aten.mul.Tensor %2161, %2160 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2166 = torch.prim.ListConstruct %int2 : (!torch.int) -> !torch.list<int>
    %2167 = torch.aten.sum.dim_IntList %2165, %2166, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2168 = torch.aten.mul.Tensor %2160, %2167 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2169 = torch.aten.sub.Tensor %2162, %2164, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2170 = torch.aten.sub.Tensor %2169, %2168, %int1 : !torch.tensor, !torch.tensor, !torch.int -> !torch.tensor
    %2171 = torch.aten.div.Tensor %result2, %2 : !torch.tensor, !torch.tensor<[],si64> -> !torch.tensor
    %2172 = torch.aten.mul.Tensor %2171, %2170 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2173 = torch.aten.mul.Tensor %2158, %2160 : !torch.tensor, !torch.tensor -> !torch.tensor
    %2174 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2175 = torch.aten.sum.dim_IntList %2173, %2174, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2176 = torch.prim.ListConstruct %int0, %int1 : (!torch.int, !torch.int) -> !torch.list<int>
    %2177 = torch.aten.sum.dim_IntList %2158, %2176, %false, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2178 = torch.prim.ListConstruct %int0 : (!torch.int) -> !torch.list<int>
    %2179 = torch.aten.sum.dim_IntList %2172, %2178, %true, %none : !torch.tensor, !torch.list<int>, !torch.bool, !torch.none -> !torch.tensor
    %2180 = torch.prim.ListConstruct %int512, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2181 = torch.aten.view %2179, %2180 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2182 = torch.prim.ListConstruct %int512, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2183 = torch.aten.new_empty %2179, %2182, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %2184 = torch.aten.zero_ %2183 : !torch.tensor -> !torch.tensor
    %2185 = torch.prim.ListConstruct %int512 : (!torch.int) -> !torch.list<int>
    %2186 = torch.aten.view %arg55, %2185 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2187 = torch.aten.ne.Scalar %2186, %int-1 : !torch.tensor, !torch.int -> !torch.tensor
    %2188 = torch.aten.unsqueeze %2187, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %2189 = torch.prim.ListConstruct %int512, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2190 = torch.aten.expand %2188, %2189, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %2191 = torch.aten.full_like %2181, %int0, %int6, %int0, %cpu, %false, %none : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.none -> !torch.tensor
    %2192 = torch.aten.where.self %2190, %2181, %2191 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %2193 = torch.prim.ListConstruct %2186 : (!torch.tensor) -> !torch.list<tensor>
    %2194 = torch.operator "aten.index_put.hacked_twin"(%2184, %2193, %2192, %true) : (!torch.tensor, !torch.list<tensor>, !torch.tensor, !torch.bool) -> !torch.tensor
    %2195 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2196 = torch.aten.view %2172, %2195 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2197 = torch.prim.ListConstruct %int2, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2198 = torch.aten.new_empty %2172, %2197, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %2199 = torch.aten.zero_ %2198 : !torch.tensor -> !torch.tensor
    %2200 = torch.aten.clone %arg54, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2201 = torch.prim.ListConstruct %int2048 : (!torch.int) -> !torch.list<int>
    %2202 = torch.aten._unsafe_view %2200, %2201 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2203 = torch.aten.ne.Scalar %2202, %int-1 : !torch.tensor, !torch.int -> !torch.tensor
    %2204 = torch.aten.unsqueeze %2203, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %2205 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2206 = torch.aten.expand %2204, %2205, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %2207 = torch.aten.full_like %2196, %int0, %int6, %int0, %cpu, %false, %none : !torch.tensor, !torch.int, !torch.int, !torch.int, !torch.Device, !torch.bool, !torch.none -> !torch.tensor
    %2208 = torch.aten.where.self %2206, %2196, %2207 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %2209 = torch.prim.ListConstruct %2202 : (!torch.tensor) -> !torch.list<tensor>
    %2210 = torch.operator "aten.index_put.hacked_twin"(%2199, %2209, %2208, %true) : (!torch.tensor, !torch.list<tensor>, !torch.tensor, !torch.bool) -> !torch.tensor
    %2211 = torch.prim.ListConstruct %int30522, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2212 = torch.aten.new_empty %2172, %2211, %int6, %int0, %cpu, %false : !torch.tensor, !torch.list<int>, !torch.int, !torch.int, !torch.Device, !torch.bool -> !torch.tensor
    %2213 = torch.aten.zero_ %2212 : !torch.tensor -> !torch.tensor
    %2214 = torch.prim.ListConstruct %int2048 : (!torch.int) -> !torch.list<int>
    %2215 = torch.aten.view %arg53, %2214 : !torch.tensor, !torch.list<int> -> !torch.tensor
    %2216 = torch.aten.ne.Scalar %2215, %int0 : !torch.tensor, !torch.int -> !torch.tensor
    %2217 = torch.aten.unsqueeze %2216, %int1 : !torch.tensor, !torch.int -> !torch.tensor
    %2218 = torch.prim.ListConstruct %int2048, %int768 : (!torch.int, !torch.int) -> !torch.list<int>
    %2219 = torch.aten.expand %2217, %2218, %false : !torch.tensor, !torch.list<int>, !torch.bool -> !torch.tensor
    %2220 = torch.aten.where.self %2219, %2196, %2207 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %2221 = torch.prim.ListConstruct %2215 : (!torch.tensor) -> !torch.list<tensor>
    %2222 = torch.aten.where.self %2219, %2196, %2207 : !torch.tensor, !torch.tensor, !torch.tensor -> !torch.tensor
    %2223 = torch.prim.TupleConstruct %2177, %2175, %2194, %2210, %2222, %2053, %2051, %2065, %2068, %2138, %2142, %2153, %2157, %2118, %2122, %2030, %2034, %2003, %2001, %2015, %2018, %1880, %1878, %1892, %1895, %1965, %1969, %1980, %1984, %1945, %1949, %1857, %1861, %1830, %1828, %1842, %1845, %323, %321, %335, %338, %408, %412, %423, %427, %388, %392, %300, %304, %273, %271, %285, %288, %150, %148, %162, %165, %235, %239, %250, %254, %215, %219, %127, %131, %100, %98, %112, %115, %1707, %1705, %1719, %1722, %1792, %1796, %1807, %1811, %1772, %1776, %1684, %1688, %1657, %1655, %1669, %1672, %1534, %1532, %1546, %1549, %1619, %1623, %1634, %1638, %1599, %1603, %1511, %1515, %1484, %1482, %1496, %1499, %1361, %1359, %1373, %1376, %1446, %1450, %1461, %1465, %1426, %1430, %1338, %1342, %1311, %1309, %1323, %1326, %1188, %1186, %1200, %1203, %1273, %1277, %1288, %1292, %1253, %1257, %1165, %1169, %1138, %1136, %1150, %1153, %1015, %1013, %1027, %1030, %1100, %1104, %1115, %1119, %1080, %1084, %992, %996, %965, %963, %977, %980, %842, %840, %854, %857, %927, %931, %942, %946, %907, %911, %819, %823, %792, %790, %804, %807, %669, %667, %681, %684, %754, %758, %769, %773, %734, %738, %646, %650, %619, %617, %631, %634, %496, %494, %508, %511, %581, %585, %596, %600, %561, %565, %473, %477, %446, %444, %458, %461, %non, %44, %47, %66, %64, %78, %81, %81, %81, %81, %81 : !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.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, tensor, tensor, tensor, tensor, tensor>
    return %2223 : !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, tensor, tensor, tensor, tensor, tensor>
  }
  torch.class_type @__torch__.torch.fx.graph_module.___torch_mangle_2.GraphModule {
    torch.method "forward", @__torch__.torch.fx.graph_module.___torch_mangle_2.GraphModule.forward
  }
  %0 = torch.nn_module {
  } : !torch.nn.Module<"__torch__.torch.fx.graph_module.___torch_mangle_2.GraphModule">
 }