File size: 11,538 Bytes
9c6594c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
import numpy
import pytest

from einops import rearrange, parse_shape, reduce
from einops.tests import is_backend_tested
from einops.tests.test_ops import imp_op_backends


def test_rearrange_examples():
    def test1(x):
        # transpose
        y = rearrange(x, "b c h w -> b h w c")
        assert tuple(y.shape) == (10, 30, 40, 20)
        return y

    def test2(x):
        # view / reshape
        y = rearrange(x, "b c h w -> b (c h w)")
        assert tuple(y.shape) == (10, 20 * 30 * 40)
        return y

    def test3(x):
        # depth-to-space
        y = rearrange(x, "b (c h1 w1) h w -> b c (h h1) (w w1)", h1=2, w1=2)
        assert tuple(y.shape) == (10, 5, 30 * 2, 40 * 2)
        return y

    def test4(x):
        # space-to-depth
        y = rearrange(x, "b c (h h1) (w w1) -> b (h1 w1 c) h w", h1=2, w1=2)
        assert tuple(y.shape) == (10, 20 * 4, 30 // 2, 40 // 2)
        return y

    def test5(x):
        # simple transposition
        y = rearrange(x, "b1 sound b2 letter -> b1 b2 sound letter")
        assert tuple(y.shape) == (10, 30, 20, 40)
        return y

    def test6(x):
        # parsing parameters
        t = rearrange(x, "b c h w -> (b h w) c")
        t = t[:, ::2]  # replacement for dot-product, just changes size of second axis
        assert tuple(t.shape) == (10 * 30 * 40, 10)

        y = rearrange(t, "(b h w) c2 -> b c2 h w", **parse_shape(x, "b _ h w"))
        assert tuple(y.shape) == (10, 10, 30, 40)
        return y

    def test7(x):
        # split of embedding into groups
        y1, y2 = rearrange(x, "b (c g) h w -> g b c h w", g=2)
        assert tuple(y1.shape) == (10, 10, 30, 40)
        assert tuple(y2.shape) == (10, 10, 30, 40)
        return y1 + y2  # only one tensor is expected in output

    def test8(x):
        # max-pooling
        y = reduce(x, "b c (h h1) (w w1) -> b c h w", reduction="max", h1=2, w1=2)
        assert tuple(y.shape) == (10, 20, 30 // 2, 40 // 2)
        return y

    def test9(x):
        # squeeze - unsqueeze
        y = reduce(x, "b c h w -> b c () ()", reduction="max")
        assert tuple(y.shape) == (10, 20, 1, 1)
        y = rearrange(y, "b c () () -> c b")
        assert tuple(y.shape) == (20, 10)
        return y

    def test10(x):
        # stack
        tensors = list(x + 0)  # 0 is needed https://github.com/tensorflow/tensorflow/issues/23185
        tensors = rearrange(tensors, "b c h w -> b h w c")
        assert tuple(tensors.shape) == (10, 30, 40, 20)
        return tensors

    def test11(x):
        # concatenate
        tensors = list(x + 0)  # 0 is needed https://github.com/tensorflow/tensorflow/issues/23185
        tensors = rearrange(tensors, "b c h w -> h (b w) c")
        assert tuple(tensors.shape) == (30, 10 * 40, 20)
        return tensors

    def shufflenet(x, convolve, c1, c2):
        # shufflenet reordering example
        x = convolve(x)
        x = rearrange(x, "b (c1 c2) h w-> b (c2 c1) h w", c1=c1, c2=c2)
        x = convolve(x)
        return x

    def convolve_strided_1d(x, stride, usual_convolution):
        x = rearrange(x, "b c t1 t2 -> b c (t1 t2)")  # reduce dimensionality
        x = rearrange(x, "b c (t stride) -> (stride b) c t", stride=stride)
        x = usual_convolution(x)
        x = rearrange(x, "(stride b) c t -> b c (t stride)", stride=stride)
        return x

    def convolve_strided_2d(x, h_stride, w_stride, usual_convolution):
        x = rearrange(x, "b c (h hs) (w ws) -> (hs ws b) c h w", hs=h_stride, ws=w_stride)
        x = usual_convolution(x)
        x = rearrange(x, "(hs ws b) c h w -> b c (h hs) (w ws)", hs=h_stride, ws=w_stride)
        return x

    def unet_like_1d(x, usual_convolution):
        # u-net like steps for increasing / reducing dimensionality
        x = rearrange(x, "b c t1 t2 -> b c (t1 t2)")  # reduce dimensionality
        y = rearrange(x, "b c (t dt) -> b (dt c) t", dt=2)
        y = usual_convolution(y)
        x = x + rearrange(y, "b (dt c) t -> b c (t dt)", dt=2)
        return x

    # mock for convolution (works for all backends)
    def convolve_mock(x):
        return x

    tests = [
        test1,
        test2,
        test3,
        test4,
        test5,
        test6,
        test7,
        test8,
        test9,
        test10,
        test11,
        lambda x: shufflenet(x, convolve=convolve_mock, c1=4, c2=5),
        lambda x: convolve_strided_1d(x, stride=2, usual_convolution=convolve_mock),
        lambda x: convolve_strided_2d(x, h_stride=2, w_stride=2, usual_convolution=convolve_mock),
        lambda x: unet_like_1d(x, usual_convolution=convolve_mock),
    ]

    for backend in imp_op_backends:
        print("testing source_examples for ", backend.framework_name)
        for test in tests:
            x = numpy.arange(10 * 20 * 30 * 40).reshape([10, 20, 30, 40])
            result1 = test(x)
            result2 = backend.to_numpy(test(backend.from_numpy(x)))
            assert numpy.array_equal(result1, result2)

            # now with strides
            x = numpy.arange(10 * 2 * 20 * 3 * 30 * 1 * 40).reshape([10 * 2, 20 * 3, 30 * 1, 40 * 1])
            # known torch bug - torch doesn't support negative steps
            last_step = -1 if (backend.framework_name != "torch" and backend.framework_name != "oneflow") else 1
            indexing_expression = numpy.index_exp[::2, ::3, ::1, ::last_step]
            result1 = test(x[indexing_expression])
            result2 = backend.to_numpy(test(backend.from_numpy(x)[indexing_expression]))
            assert numpy.array_equal(result1, result2)


def tensor_train_example_numpy():
    # kept here just for a collection, only tested for numpy
    # https://arxiv.org/pdf/1509.06569.pdf, (5)
    x = numpy.ones([3, 4, 5, 6])
    rank = 4
    if numpy.__version__ < "1.15.0":
        # numpy.einsum fails here, skip test
        return
    # creating appropriate Gs
    Gs = [numpy.ones([d, d, rank, rank]) for d in x.shape]
    Gs[0] = Gs[0][:, :, :1, :]
    Gs[-1] = Gs[-1][:, :, :, :1]

    # einsum way
    y = x.reshape((1,) + x.shape)
    for G in Gs:
        # taking partial results left-to-right
        # y = numpy.einsum('i j alpha beta, alpha i ...  -> beta ... j', G, y)
        y = numpy.einsum("i j a b, a i ...  -> b ... j", G, y)
    y1 = y.reshape(-1)

    # alternative way
    y = x.reshape(-1)
    for G in Gs:
        i, j, alpha, beta = G.shape
        y = rearrange(y, "(i rest alpha) -> rest (alpha i)", alpha=alpha, i=i)
        y = y @ rearrange(G, "i j alpha beta -> (alpha i) (j beta)")
        y = rearrange(y, "rest (beta j) -> (beta rest j)", beta=beta, j=j)
    y2 = y
    assert numpy.allclose(y1, y2)

    # yet another way
    y = x
    for G in Gs:
        i, j, alpha, beta = G.shape
        y = rearrange(y, "i ... (j alpha) -> ... j (alpha i)", alpha=alpha, i=i)
        y = y @ rearrange(G, "i j alpha beta -> (alpha i) (j beta)")
    y3 = y.reshape(-1)
    assert numpy.allclose(y1, y3)


def test_pytorch_yolo_fragment():
    if not is_backend_tested("torch"):
        pytest.skip()

    import torch

    def old_way(input, num_classes, num_anchors, anchors, stride_h, stride_w):
        # https://github.com/BobLiu20/YOLOv3_PyTorch/blob/c6b483743598b5f64d520d81e7e5f47ba936d4c9/nets/yolo_loss.py#L28-L44
        bs = input.size(0)
        in_h = input.size(2)
        in_w = input.size(3)
        scaled_anchors = [(a_w / stride_w, a_h / stride_h) for a_w, a_h in anchors]

        prediction = input.view(bs, num_anchors, 5 + num_classes, in_h, in_w).permute(0, 1, 3, 4, 2).contiguous()
        # Get outputs
        x = torch.sigmoid(prediction[..., 0])  # Center x
        y = torch.sigmoid(prediction[..., 1])  # Center y
        w = prediction[..., 2]  # Width
        h = prediction[..., 3]  # Height
        conf = torch.sigmoid(prediction[..., 4])  # Conf
        pred_cls = torch.sigmoid(prediction[..., 5:])  # Cls pred.

        # https://github.com/BobLiu20/YOLOv3_PyTorch/blob/c6b483743598b5f64d520d81e7e5f47ba936d4c9/nets/yolo_loss.py#L70-L92
        FloatTensor = torch.cuda.FloatTensor if x.is_cuda else torch.FloatTensor
        LongTensor = torch.cuda.LongTensor if x.is_cuda else torch.LongTensor
        # Calculate offsets for each grid
        grid_x = (
            torch.linspace(0, in_w - 1, in_w)
            .repeat(in_w, 1)
            .repeat(bs * num_anchors, 1, 1)
            .view(x.shape)
            .type(FloatTensor)
        )
        grid_y = (
            torch.linspace(0, in_h - 1, in_h)
            .repeat(in_h, 1)
            .t()
            .repeat(bs * num_anchors, 1, 1)
            .view(y.shape)
            .type(FloatTensor)
        )
        # Calculate anchor w, h
        anchor_w = FloatTensor(scaled_anchors).index_select(1, LongTensor([0]))
        anchor_h = FloatTensor(scaled_anchors).index_select(1, LongTensor([1]))
        anchor_w = anchor_w.repeat(bs, 1).repeat(1, 1, in_h * in_w).view(w.shape)
        anchor_h = anchor_h.repeat(bs, 1).repeat(1, 1, in_h * in_w).view(h.shape)
        # Add offset and scale with anchors
        pred_boxes = FloatTensor(prediction[..., :4].shape)
        pred_boxes[..., 0] = x.data + grid_x
        pred_boxes[..., 1] = y.data + grid_y
        pred_boxes[..., 2] = torch.exp(w.data) * anchor_w
        pred_boxes[..., 3] = torch.exp(h.data) * anchor_h
        # Results
        _scale = torch.Tensor([stride_w, stride_h] * 2).type(FloatTensor)
        output = torch.cat(
            (pred_boxes.view(bs, -1, 4) * _scale, conf.view(bs, -1, 1), pred_cls.view(bs, -1, num_classes)), -1
        )
        return output

    def new_way(input, num_classes, num_anchors, anchors, stride_h, stride_w):
        raw_predictions = rearrange(input, " b (anchor prediction) h w -> prediction b anchor h w", anchor=num_anchors)

        anchors = torch.FloatTensor(anchors).to(input.device)
        anchor_sizes = rearrange(anchors, "anchor dim -> dim () anchor () ()")

        _, _, _, in_h, in_w = raw_predictions.shape
        grid_h = rearrange(torch.arange(in_h).float(), "h -> () () h ()").to(input.device)
        grid_w = rearrange(torch.arange(in_w).float(), "w -> () () () w").to(input.device)

        predicted_bboxes = torch.zeros_like(raw_predictions)
        predicted_bboxes[0] = (raw_predictions[0].sigmoid() + grid_h) * stride_h  # center y
        predicted_bboxes[1] = (raw_predictions[1].sigmoid() + grid_w) * stride_w  # center x
        predicted_bboxes[2:4] = (raw_predictions[2:4].exp()) * anchor_sizes  # bbox width and height
        predicted_bboxes[4] = raw_predictions[4].sigmoid()  # confidence
        predicted_bboxes[5:] = raw_predictions[5:].sigmoid()  # class predictions
        # only to match results of original code, not needed
        return rearrange(predicted_bboxes, "prediction b anchor h w -> b anchor h w prediction")

    stride_h = 4
    stride_w = 4
    batch_size = 5
    num_classes = 12
    anchors = [[50, 100], [100, 50], [75, 75]]
    num_anchors = len(anchors)

    input = torch.randn([batch_size, num_anchors * (5 + num_classes), 1, 1])
    result1 = old_way(
        input=input,
        num_anchors=num_anchors,
        num_classes=num_classes,
        stride_h=stride_h,
        stride_w=stride_w,
        anchors=anchors,
    )
    result2 = new_way(
        input=input,
        num_anchors=num_anchors,
        num_classes=num_classes,
        stride_h=stride_h,
        stride_w=stride_w,
        anchors=anchors,
    )
    result1 = result1.reshape(result2.shape)
    assert torch.allclose(result1, result2)