From 977b3ff3ada203a4d990749e79c354f19f8d8473 Mon Sep 17 00:00:00 2001 From: Ina_Dobreva Date: Tue, 29 Oct 2019 17:36:06 +0000 Subject: [PATCH 1/2] Add test for the qnn_add operator The tests use fake quant approach so until the tf session tensors remain in float32. The test data has to be passed in uint8 because of how the tflite/tvm comparison works. Abs tolerance up to 1 is allowed for the qnn results. For now input_stats are hardcoded assuming the tests for the other qnn ops will pass the input data in the same range. --- tests/python/frontend/tflite/test_forward.py | 87 +++++++++++++++----- 1 file changed, 65 insertions(+), 22 deletions(-) diff --git a/tests/python/frontend/tflite/test_forward.py b/tests/python/frontend/tflite/test_forward.py index de19fe34f811..fc707c8294c8 100644 --- a/tests/python/frontend/tflite/test_forward.py +++ b/tests/python/frontend/tflite/test_forward.py @@ -122,7 +122,7 @@ def run_tflite_graph(tflite_model_buf, input_data): def compare_tflite_with_tvm(in_data, in_name, input_tensors, - output_tensors, init_global_variables=False, out_names=None): + output_tensors, init_global_variables=False, out_names=None, quantized=False): """Generic function to generate and compare TFLite and TVM output""" in_data = convert_to_list(in_data) in_name = convert_to_list(in_name) @@ -137,6 +137,17 @@ def compare_tflite_with_tvm(in_data, in_name, input_tensors, # convert to tflite model converter = interpreter_wrapper.TFLiteConverter.from_session( sess, input_tensors, output_tensors) + + if quantized: + converter.inference_type = tf.lite.constants.QUANTIZED_UINT8 + input_arrays = converter.get_input_arrays() + input_stats = {} + # hardcode the mean_values and std_dev_values (m,s) to be the same for all inputs + # s = 255/(fmax-fmin); m = -fmin*s (the zero point) + for i in input_arrays: + input_stats[i] = (128., 1.275) + converter.quantized_input_stats = input_stats + tflite_model_buffer = converter.convert() tflite_output = run_tflite_graph(tflite_model_buffer, in_data) @@ -148,8 +159,13 @@ def compare_tflite_with_tvm(in_data, in_name, input_tensors, tvm_output = run_tvm_graph(tflite_model_buffer, in_data, in_node, target=device, num_output=len(out_names), out_names=out_names) - for i in range(len(tflite_output)): - tvm.testing.assert_allclose(tflite_output[i], tvm_output[i], atol=1e-5, rtol=1e-5) + if quantized: + for i in range(len(tflite_output)): + # allow absolute tolerance of 1 in the quantized results + tvm.testing.assert_allclose(tflite_output[i], tvm_output[i], atol=1, rtol=1e-5) + else: + for i in range(len(tflite_output)): + tvm.testing.assert_allclose(tflite_output[i], tvm_output[i], atol=1e-5, rtol=1e-5) def with_fused_activation_function(input_tensor, fn_name): @@ -545,34 +561,55 @@ def test_forward_concatenation(): # Element-wise # --- -def _test_elemwise(math_op, data, fused_activation_function=None): +def _test_elemwise(math_op, data, fused_activation_function=None, quantized=False): """ One iteration of elemwise """ assert len(data) == 2 # Test with two tensors with tf.Graph().as_default(): - in_data = [array_ops.placeholder(shape=data[0].shape, dtype=data[0].dtype, name='in_0'), - array_ops.placeholder(shape=data[1].shape, dtype=data[1].dtype, name='in_1')] - out = math_op(in_data[0], in_data[1]) - out = with_fused_activation_function(out, fused_activation_function) - compare_tflite_with_tvm(data, ['in_0:0', 'in_1:0'], in_data, [out]) + in_data = [array_ops.placeholder(shape=data[0].shape, dtype='float32', name='in_0'), + array_ops.placeholder(shape=data[1].shape, dtype='float32', name='in_1')] + + if quantized: + # fake_quant will keep the tensors in float32 until the conversion in the session + inq_data = [tf.quantization.fake_quant_with_min_max_args(in_data[0], min=-100, max=100, name="inq_0"), + tf.quantization.fake_quant_with_min_max_args(in_data[1], min=-100, max=100, name="inq_1")] + out = math_op(inq_data[0], inq_data[1]) + out = with_fused_activation_function(out, fused_activation_function) + out = tf.quantization.fake_quant_with_min_max_args(out, min=-200, max=200, name="out") + compare_tflite_with_tvm(data, ['inq_0:0', 'inq_1:0'], inq_data, [out], quantized=True) + + else: + out = math_op(in_data[0], in_data[1]) + out = with_fused_activation_function(out, fused_activation_function) + compare_tflite_with_tvm(data, ['in_0:0', 'in_1:0'], in_data, [out]) # Test with tensor and constant with tf.Graph().as_default(): - in_data = [array_ops.placeholder(shape=data[0].shape, dtype=data[0].dtype, name='in')] - out = math_op(in_data[0], ops.convert_to_tensor(data[1], dtype=data[1].dtype)) - out = with_fused_activation_function(out, fused_activation_function) - compare_tflite_with_tvm([data[0]], ['in:0'], in_data, [out]) + in_data = [array_ops.placeholder(shape=data[0].shape, dtype='float32', name='in_0')] + if quantized: + inq_data = [tf.quantization.fake_quant_with_min_max_args(in_data[0], min=-100, max=100, name="inq_0")] + inq_const = tf.quantization.fake_quant_with_min_max_args(data[1], min=-100, max=100, name="const_tensor") + # the 2nd tensor is treated as constant and directly added as part of the operation + out = math_op(inq_data, ops.convert_to_tensor(inq_const, dtype='float32', name='inq_const')) + out = with_fused_activation_function(out, fused_activation_function) + out = tf.quantization.fake_quant_with_min_max_args(out, min=-200, max=200, name="out") + compare_tflite_with_tvm(data[0], ['inq_0:0'], inq_data, [out], quantized=True) + + else: + out = math_op(in_data[0], ops.convert_to_tensor(data[1], dtype=data[1].dtype)) + out = with_fused_activation_function(out, fused_activation_function) + compare_tflite_with_tvm(data[0], ['in_0:0'], in_data, [out]) ####################################################################### # Add # --- -def _test_add(data, fused_activation_function=None): +def _test_add(data, fused_activation_function=None, quantized=False): """ One iteration of add """ - return _test_elemwise(math_ops.add, data, fused_activation_function) + return _test_elemwise(math_ops.add, data, fused_activation_function, quantized) ####################################################################### # Subtract @@ -624,17 +661,23 @@ def _test_greater(data): """ One iteration of greater """ return _test_elemwise(math_ops.greater, data) -def _test_forward_elemwise(testop): +def _test_forward_elemwise(testop, quantized=False): """ Elewise""" - testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 1, 3)), - np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 1, 3))]) - testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 3)), - np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 3))]) - testop([np.arange(3.0, dtype=np.float32).reshape((1, 3)), - np.arange(1.0, 4.0, dtype=np.float32).reshape((1, 3))]) + # test with two quantized tensors + if quantized: + testop([np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8), + np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8)], quantized=True) + else: + testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 1, 3)), + np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 1, 3))]) + testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 3)), + np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 3))]) + testop([np.arange(3.0, dtype=np.float32).reshape((1, 3)), + np.arange(1.0, 4.0, dtype=np.float32).reshape((1, 3))]) def test_all_elemwise(): _test_forward_elemwise(_test_add) + _test_forward_elemwise(_test_add, quantized=True) _test_forward_elemwise(partial(_test_add, fused_activation_function="RELU")) _test_forward_elemwise(partial(_test_add, fused_activation_function="RELU6")) _test_forward_elemwise(_test_sub) From 636e640adb9fc41d28808c1bd90fd5145dcfd0fd Mon Sep 17 00:00:00 2001 From: Ina_Dobreva Date: Mon, 11 Nov 2019 10:16:31 +0000 Subject: [PATCH 2/2] Separate qnn uint8 test function from the fp32 elemwise tests Isolate qnn uint8 elemwise tests Remove blank lines --- tests/python/frontend/tflite/test_forward.py | 27 +++++++++----------- 1 file changed, 12 insertions(+), 15 deletions(-) diff --git a/tests/python/frontend/tflite/test_forward.py b/tests/python/frontend/tflite/test_forward.py index fc707c8294c8..2612cfd8be18 100644 --- a/tests/python/frontend/tflite/test_forward.py +++ b/tests/python/frontend/tflite/test_forward.py @@ -579,7 +579,6 @@ def _test_elemwise(math_op, data, fused_activation_function=None, quantized=Fals out = with_fused_activation_function(out, fused_activation_function) out = tf.quantization.fake_quant_with_min_max_args(out, min=-200, max=200, name="out") compare_tflite_with_tvm(data, ['inq_0:0', 'inq_1:0'], inq_data, [out], quantized=True) - else: out = math_op(in_data[0], in_data[1]) out = with_fused_activation_function(out, fused_activation_function) @@ -597,7 +596,6 @@ def _test_elemwise(math_op, data, fused_activation_function=None, quantized=Fals out = with_fused_activation_function(out, fused_activation_function) out = tf.quantization.fake_quant_with_min_max_args(out, min=-200, max=200, name="out") compare_tflite_with_tvm(data[0], ['inq_0:0'], inq_data, [out], quantized=True) - else: out = math_op(in_data[0], ops.convert_to_tensor(data[1], dtype=data[1].dtype)) out = with_fused_activation_function(out, fused_activation_function) @@ -661,23 +659,22 @@ def _test_greater(data): """ One iteration of greater """ return _test_elemwise(math_ops.greater, data) -def _test_forward_elemwise(testop, quantized=False): +def _test_forward_elemwise(testop): """ Elewise""" - # test with two quantized tensors - if quantized: - testop([np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8), - np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8)], quantized=True) - else: - testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 1, 3)), - np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 1, 3))]) - testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 3)), - np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 3))]) - testop([np.arange(3.0, dtype=np.float32).reshape((1, 3)), - np.arange(1.0, 4.0, dtype=np.float32).reshape((1, 3))]) + testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 1, 3)), + np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 1, 3))]) + testop([np.arange(6.0, dtype=np.float32).reshape((2, 1, 3)), + np.arange(1.0, 7.0, dtype=np.float32).reshape((2, 1, 3))]) + testop([np.arange(3.0, dtype=np.float32).reshape((1, 3)), + np.arange(1.0, 4.0, dtype=np.float32).reshape((1, 3))]) + +def _test_forward_elemwise_quantized(testop): + testop([np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8), + np.array(np.random.uniform(0, 255, (3, 6)), dtype=np.uint8)], quantized=True) def test_all_elemwise(): _test_forward_elemwise(_test_add) - _test_forward_elemwise(_test_add, quantized=True) + _test_forward_elemwise_quantized(_test_add) _test_forward_elemwise(partial(_test_add, fused_activation_function="RELU")) _test_forward_elemwise(partial(_test_add, fused_activation_function="RELU6")) _test_forward_elemwise(_test_sub)