const auto float_graph = model_->subgraphs()->Get(subgraph_idx);
    ASSERT_EQ(quantized_graph->tensors()->size(),
              float_graph->tensors()->size());
    // Make sure the graph only has one Conv operation.
    ASSERT_EQ(quantized_graph->operators()->size(), 1);
    const auto op = quantized_graph->operators()->Get(0);
    const uint32_t op_code_idx = op->opcode_index();

				}
				if ft.Size() == 8 {
					return s.float64ToUint32(n, v, ft, tt)
				}
			}
		}

		if !ok {
			s.Fatalf("weird float conversion %v -> %v", ft, tt)
		}
		op1, op2, it := conv.op1, conv.op2, conv.intermediateType

		if op1 != ssa.OpInvalid && op2 != ssa.OpInvalid {
			// normal case, not tripping over unsigned 64
			if op1 == ssa.OpCopy {
				if op2 == ssa.OpCopy {
					return v
				}

};

using ConvertConv2DDynamic =
    ConvertConvDynamic<TF::Conv2DOp, /*num_spatial_dims=*/2>;

// Converts the TensorFlow conv op in template to the generic HLO conv op by
// converting TensorFlow op attributes to HLO op attributes.
//
// Sample result for Conv2D:
//
//   %conv = "mhlo.convolution"(%input, %filter) {
//     strides = [1, 2],
//     paddings = [[1, 0], [1, 1]],
//     ...
//   }
//

      }
      func @_func(%input: tensor<2x112x112x12xf32>, %filter: tensor<7x7x3x64xf32>) {
        %filter_transform = "tf.Pad/tf.Transpose/tf.Reshape"(%filter): tensor<7x7x3x64xf32>) -> tensor<4x4x12x64xf32>
        %conv = "tf.Conv2D"(%input, %filter_transfrom) {strides = [1, 1, 1, 1]}: (tensor<2x112x112x12xf32>, tensor<4x4x12x64xf32>) -> tensor<2x112x112x64xf32>
      }
    }
    ```

used for communication with all hosts.}]>:$network_configs
  );

  let results = (outs);

  TF_DerivedOperandSizeAttr N = TF_DerivedOperandSizeAttr<0>;
}

def TF_ConvOp : TF_Op<"Conv", [Pure]> {
  let summary = [{
Computes a N-D convolution given (N+1+batch_dims)-D `input` and (N+2)-D `filter` tensors.
  }];

  let description = [{