[QNN] Concatenate operator (apache#3730)

python/tvm/relay/qnn/op/qnn.py

@@ -18,6 +18,7 @@
 """QNN dialect operators."""
 
 from __future__ import absolute_import as _abs
+from tvm import relay
 from . import _make
 
 def requantize(data,
@@ -72,3 +73,75 @@ def requantize(data,
                             output_zero_point,
                             rounding,
                             out_dtype)
+
+def concatenate(data,
+                input_scales,
+                input_zero_points,
+                output_scale,
+                output_zero_point,
+                axis):
+    """Concatenate the quantized input tensors along the given axis.
+
+    Parameters
+    ----------
+    data : Union(List[relay.Expr], Tuple[relay.Expr])
+        The list of quantized tensors.
+
+    input_scales : List[float32]
+        The list of scales of input quantized tensors.
+
+    input_zero_points : List[int32]
+        The list of zero points of input quantized tensors.
+
+    output_scale : float32
+        The scale of the output quantized tensor.
+
+    output_zero_point : int32
+        The zero point of the output quantized tensor.
+
+    axis : int
+        The axis along which the tensors are concatenated.
+
+    Returns
+    -------
+    result: relay.Expr
+        The concatenated quantized tensor.
+    """
+
+    data = list(data)
+    requantized_exprs = list(data)
+
+    # Find the dtype of the input expr. This is required for the requantize op. Since, this is
+    # concatenate op, the dtype of the input is same as dtype of the output.
+    data0 = relay.transform.infer_type(data[0])
+    in_dtype = data0.checked_type.dtype
+
+    # First check if all the input qnn params match. If yes, we can call concatenate first, followed
+    # by a requantize.
+    if all(scale == input_scales[0] for scale in input_scales)\
+            and all(zero_point == input_zero_points[0] for zero_point in input_zero_points):
+        out = relay.concatenate(tuple(data), axis)
+        input_scale = input_scales[0]
+        input_zero_point = input_zero_points[0]
+        if input_scale != output_scale or input_zero_point != output_zero_point:
+            out = requantize(data=out,
+                             input_scale=input_scales[0],
+                             input_zero_point=input_zero_points[0],
+                             output_scale=output_scale,
+                             output_zero_point=output_zero_point,
+                             out_dtype=in_dtype)
+        return out
+
+    # If the output qnn params do not match the input qnn params, we can call requantize on the
+    # input expr first, followed by a concatenate on the requantized input exprs.
+    for idx, quantized_expr in enumerate(data):
+        input_scale = input_scales[idx]
+        input_zero_point = input_zero_points[idx]
+        if input_scale != output_scale or input_zero_point != output_zero_point:
+            requantized_exprs[idx] = requantize(data=quantized_expr,
+                                                input_scale=input_scale,
+                                                input_zero_point=input_zero_point,
+                                                output_scale=output_scale,
+                                                output_zero_point=output_zero_point,
+                                                out_dtype=in_dtype)
+    return relay.concatenate(tuple(requantized_exprs), axis)

tests/python/relay/test_qnn_concatenate.py

@@ -0,0 +1,145 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+import tvm
+import numpy as np
+from tvm import relay
+from tvm.contrib import graph_runtime
+import topi.testing
+
+def test_same_io_qnn_params():
+    data_dtype = 'int32'
+    axis = 0
+    x_data = np.arange(-32, 32, 1).reshape(1, 64).astype(data_dtype)
+    y_data = np.arange(-64, 64, 2).reshape(1, 64).astype(data_dtype)
+    x_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+    y_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+
+    x = relay.var("x", shape=(1, 64), dtype=data_dtype)
+    y = relay.var("y", shape=(1, 64), dtype=data_dtype)
+    z = relay.qnn.op.concatenate((x, y),
+                                 input_scales=[x_scale, y_scale],
+                                 input_zero_points=[0, 0],
+                                 output_scale=y_scale,
+                                 output_zero_point=0,
+                                 axis=axis)
+
+    func = relay.Function([x, y], z)
+    assert func.astext().count('requantize') == 0
+    mod = relay.Module.from_expr(func)
+    mod = relay.transform.Legalize()(mod)
+    func = mod["main"]
+
+    golden_output = np.concatenate((x_data, y_data), axis=axis)
+
+    intrp = relay.create_executor("graph", ctx=tvm.cpu(0), target="llvm")
+    op_res = intrp.evaluate(func)(x_data, y_data)
+    np.testing.assert_equal(op_res.asnumpy(), golden_output)
+
+def test_different_io_qnn_params():
+    data_dtype = 'int32'
+    axis = 0
+    x_data = np.arange(-32, 32, 1).reshape(1, 64).astype(data_dtype)
+    y_data = np.arange(-64, 64, 2).reshape(1, 64).astype(data_dtype)
+    x_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+    y_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+
+    x = relay.var("x", shape=(1, 64), dtype=data_dtype)
+    y = relay.var("y", shape=(1, 64), dtype=data_dtype)
+    z = relay.qnn.op.concatenate((x, y),
+                                 input_scales=[x_scale, y_scale],
+                                 input_zero_points=[3, 4],
+                                 output_scale=y_scale,
+                                 output_zero_point=1,
+                                 axis=axis)
+
+    func = relay.Function([x, y], z)
+    assert func.astext().count('requantize') == 2
+    mod = relay.Module.from_expr(func)
+    mod = relay.transform.Legalize()(mod)
+    func = mod["main"]
+
+    golden_output = np.concatenate((x_data - 2, y_data - 3), axis=axis)
+
+    intrp = relay.create_executor("graph", ctx=tvm.cpu(0), target="llvm")
+    op_res = intrp.evaluate(func)(x_data, y_data)
+    np.testing.assert_equal(op_res.asnumpy(), golden_output)
+
+def test_few_same_io_qnn_params():
+    data_dtype = 'int32'
+    axis = 0
+    x_data = np.arange(-32, 32, 1).reshape(1, 64).astype(data_dtype)
+    y_data = np.arange(-64, 64, 2).reshape(1, 64).astype(data_dtype)
+    x_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+    y_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+
+    x = relay.var("x", shape=(1, 64), dtype=data_dtype)
+    y = relay.var("y", shape=(1, 64), dtype=data_dtype)
+    z = relay.qnn.op.concatenate((x, y),
+                                 input_scales=[x_scale, y_scale],
+                                 input_zero_points=[0, 1],
+                                 output_scale=y_scale,
+                                 output_zero_point=1,
+                                 axis=axis)
+
+    func = relay.Function([x, y], z)
+    assert func.astext().count('requantize') == 1
+    mod = relay.Module.from_expr(func)
+    mod = relay.transform.Legalize()(mod)
+    func = mod["main"]
+
+    golden_output = np.concatenate((x_data + 1, y_data), axis=axis)
+
+    intrp = relay.create_executor("graph", ctx=tvm.cpu(0), target="llvm")
+    op_res = intrp.evaluate(func)(x_data, y_data)
+    np.testing.assert_equal(op_res.asnumpy(), golden_output)
+
+def test_same_i_qnn_params():
+    data_dtype = 'int32'
+    axis = 0
+    x_data = np.arange(-32, 32, 1).reshape(1, 64).astype(data_dtype)
+    y_data = np.arange(-64, 64, 2).reshape(1, 64).astype(data_dtype)
+    x_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+    y_scale = (62 + 64) / (np.power(2, 32) - 1.0)
+
+    x = relay.var("x", shape=(1, 64), dtype=data_dtype)
+    y = relay.var("y", shape=(1, 64), dtype=data_dtype)
+    z = relay.qnn.op.concatenate((x, y),
+                                 input_scales=[x_scale, y_scale],
+                                 input_zero_points=[0, 0],
+                                 output_scale=y_scale,
+                                 output_zero_point=1,
+                                 axis=axis)
+
+    func = relay.Function([x, y], z)
+    assert func.astext().count('requantize') == 1
+    mod = relay.Module.from_expr(func)
+    mod = relay.transform.Legalize()(mod)
+    func = mod["main"]
+
+    golden_output = np.concatenate((x_data + 1, y_data + 1), axis=axis)
+
+    intrp = relay.create_executor("graph", ctx=tvm.cpu(0), target="llvm")
+    op_res = intrp.evaluate(func)(x_data, y_data)
+    np.testing.assert_equal(op_res.asnumpy(), golden_output)
+
+
+if __name__ == '__main__':
+    test_same_io_qnn_params()
+    test_different_io_qnn_params()
+    test_few_same_io_qnn_params()
+    test_same_i_qnn_params()