onnx.cpp 52.8 KB
Newer Older
Paul's avatar
Paul committed
1
2
3
4
5
6
7
8
#include <google/protobuf/text_format.h>
#include <google/protobuf/io/zero_copy_stream_impl.h>
#include <onnx.pb.h>
#include <iostream>
#include <fstream>
#include <unordered_map>
#include <functional>
#include <array>
Paul's avatar
Paul committed
9
#include <utility>
10
#include <vector>
Paul's avatar
Paul committed
11

Paul's avatar
Paul committed
12
13
14
15
16
17
#include <migraphx/fallthrough.hpp>
#include <migraphx/program.hpp>
#include <migraphx/operators.hpp>
#include <migraphx/ranges.hpp>
#include <migraphx/instruction.hpp>
#include <migraphx/config.hpp>
18
#include <migraphx/onnx.hpp>
Paul's avatar
Paul committed
19
20

namespace migraphx {
Paul's avatar
Paul committed
21
inline namespace MIGRAPHX_INLINE_NS {
Paul's avatar
Paul committed
22
23
24
25
26

struct onnx_parser
{
    using attribute_map = std::unordered_map<std::string, onnx::AttributeProto>;
    using node_map      = std::unordered_map<std::string, onnx::NodeProto>;
Paul's avatar
Paul committed
27
28
    using op_func =
        std::function<std::vector<instruction_ref>(attribute_map, std::vector<instruction_ref>)>;
Paul's avatar
Paul committed
29
30
    node_map nodes;
    std::unordered_map<std::string, instruction_ref> instructions;
Scott Thornton's avatar
Scott Thornton committed
31
    program prog    = program();
32
    bool is_pytorch = false;
Paul's avatar
Paul committed
33
34

    std::unordered_map<std::string, op_func> ops;
35
    std::unordered_map<std::string, operation> map_actv_funcs;
Paul's avatar
Paul committed
36
37
38

    onnx_parser()
    {
Shucai Xiao's avatar
Shucai Xiao committed
39
        add_generic_op("MatMul", op::dot{});
Khalique's avatar
Khalique committed
40
        add_generic_op("Relu", op::relu{});
Khalique's avatar
Khalique committed
41
42
        add_generic_op("Sigmoid", op::sigmoid{});
        add_generic_op("Abs", op::abs{});
Shucai Xiao's avatar
Shucai Xiao committed
43
44
        add_generic_op("Exp", op::exp{});
        add_generic_op("Log", op::log{});
Khalique's avatar
Khalique committed
45
46
        // disable dropout for inference
        add_generic_op("Dropout", op::identity{});
Khalique's avatar
Khalique committed
47
        add_generic_op("Identity", op::identity{});
Shucai Xiao's avatar
Shucai Xiao committed
48
49
50
        add_generic_op("Sin", op::sin{});
        add_generic_op("Cos", op::cos{});
        add_generic_op("Tan", op::tan{});
51
52
        add_generic_op("Sinh", op::sinh{});
        add_generic_op("Cosh", op::cosh{});
53
        add_generic_op("Tanh", op::tanh{});
54
55
56
        add_generic_op("Asin", op::asin{});
        add_generic_op("Acos", op::acos{});
        add_generic_op("Atan", op::atan{});
Paul's avatar
Paul committed
57

Khalique's avatar
Khalique committed
58
59
60
61
62
        add_binary_op("Add", op::add{});
        add_binary_op("Div", op::div{});
        add_binary_op("Mul", op::mul{});
        add_binary_op("Sub", op::sub{});

Khalique's avatar
Khalique committed
63
64
65
        add_variadic_op("Sum", op::add{});
        add_variadic_op("Max", op::max{});
        add_variadic_op("Min", op::min{});
Paul's avatar
Paul committed
66

Khalique's avatar
Khalique committed
67
        add_mem_op("LRN", &onnx_parser::parse_lrn);
Khalique's avatar
Khalique committed
68
        add_mem_op("ImageScaler", &onnx_parser::parse_imagescaler);
69
        add_mem_op("LeakyRelu", &onnx_parser::parse_leaky_relu);
Khalique's avatar
Khalique committed
70
        add_mem_op("Elu", &onnx_parser::parse_elu);
Paul's avatar
Paul committed
71
72
        add_mem_op("Constant", &onnx_parser::parse_constant);
        add_mem_op("Conv", &onnx_parser::parse_conv);
Paul's avatar
Paul committed
73
74
        add_mem_op("MaxPool", &onnx_parser::parse_pooling);
        add_mem_op("AveragePool", &onnx_parser::parse_pooling);
75
76
        add_mem_op("GlobalMaxPool", &onnx_parser::parse_pooling);
        add_mem_op("GlobalAveragePool", &onnx_parser::parse_pooling);
Paul's avatar
Paul committed
77
        add_mem_op("Reshape", &onnx_parser::parse_reshape);
Paul's avatar
Paul committed
78
79
        add_mem_op("Flatten", &onnx_parser::parse_flatten);
        add_mem_op("Gemm", &onnx_parser::parse_gemm);
80
        add_mem_op("BatchNormalization", &onnx_parser::parse_batchnorm);
Paul's avatar
Paul committed
81
        add_mem_op("Softmax", &onnx_parser::parse_softmax);
82
83
84
        add_mem_op("Squeeze", &onnx_parser::parse_squeeze);
        add_mem_op("Unsqueeze", &onnx_parser::parse_unsqueeze);
        add_mem_op("Slice", &onnx_parser::parse_slice);
Scott Thornton's avatar
Scott Thornton committed
85
        add_mem_op("Concat", &onnx_parser::parse_concat);
86
87
88
        add_mem_op("Gather", &onnx_parser::parse_gather);
        add_mem_op("Shape", &onnx_parser::parse_shape);
        add_mem_op("ConstantFill", &onnx_parser::parse_constant_fill);
Khalique's avatar
Khalique committed
89
        add_mem_op("Transpose", &onnx_parser::parse_transpose);
Shucai Xiao's avatar
Shucai Xiao committed
90
        add_mem_op("RNN", &onnx_parser::parse_rnn);
91
        add_mem_op("GRU", &onnx_parser::parse_gru);
Shucai Xiao's avatar
Shucai Xiao committed
92
        add_mem_op("LSTM", &onnx_parser::parse_lstm);
Khalique's avatar
Khalique committed
93
        add_mem_op("Pad", &onnx_parser::parse_pad);
94
95
96
97
98
99
100

        // init the activation function map
        init_actv_func();
    }

    void init_actv_func()
    {
101
102
103
104
105
        map_actv_funcs.insert(std::make_pair("tanh", op::tanh{}));
        map_actv_funcs.insert(std::make_pair("relu", op::relu{}));
        map_actv_funcs.insert(std::make_pair("sigmoid", op::sigmoid{}));
        map_actv_funcs.insert(std::make_pair("leakyrelu", op::leaky_relu{}));
        map_actv_funcs.insert(std::make_pair("elu", op::elu{}));
Paul's avatar
Paul committed
106
107
108
109
    }

    template <class F>
    void add_op(std::string name, F f)
Paul's avatar
Paul committed
110
111
112
113
114
115
116
117
118
    {
        ops.emplace(name, [=](auto&&... xs) {
            return std::vector<instruction_ref>{f(std::forward<decltype(xs)>(xs)...)};
        });
    }

    // Multi output op
    template <class F>
    void add_multi_op(std::string name, F f)
Paul's avatar
Paul committed
119
120
121
122
123
124
125
    {
        ops.emplace(name, f);
    }

    template <class F>
    void add_mem_op(std::string name, F f)
    {
Paul's avatar
Paul committed
126
        add_op(name, [=](auto&&... xs) {
Paul's avatar
Paul committed
127
128
129
            return std::mem_fn(f)(*this, name, std::forward<decltype(xs)>(xs)...);
        });
    }
Khalique's avatar
Khalique committed
130

131
    template <class T>
Khalique's avatar
Khalique committed
132
    void add_binary_op(std::string name, T x)
133
    {
Paul's avatar
Paul committed
134
        add_op(name, [this, x](attribute_map attributes, std::vector<instruction_ref> args) {
Scott Thornton's avatar
Scott Thornton committed
135
            if(args.size() != 2)
Paul's avatar
Paul committed
136
                MIGRAPHX_THROW("binary operators should have 2 operands");
137
            if(contains(attributes, "broadcast") and contains(attributes, "axis"))
138
139
140
141
            {
                uint64_t broadcasted = parse_value(attributes.at("broadcast")).at<uint64_t>();
                if(broadcasted != 0)
                {
142
                    uint64_t axis = parse_value(attributes.at("axis")).at<uint64_t>();
143
144
145
146
                    auto l =
                        prog.add_instruction(op::broadcast{axis, args[0]->get_shape()}, args[1]);
                    return prog.add_instruction(x, args[0], l);
                }
147
                return prog.add_instruction(x, args);
148
            }
Paul's avatar
Paul committed
149
            else
150
            {
Khalique's avatar
Khalique committed
151
                return add_broadcastable_binary_op(args[0], args[1], x);
152
153
154
155
            }
        });
    }

Khalique's avatar
Khalique committed
156
157
158
159
160
    template <class T>
    instruction_ref add_broadcastable_binary_op(instruction_ref arg0, instruction_ref arg1, T x)
    {
        if(arg0->get_shape() != arg1->get_shape())
        {
Khalique's avatar
Khalique committed
161
162
163
164
165
166
167
168
169
170
171
172
173
            // Example:
            // s0 = (3,2,4,5) and s1 = (2,1,1)
            //
            // In this case we need to broadcast (:,1,1) portion of
            // s1 plus broadcast the 1st dimension of s1
            // giving output_lens = (3,2,4,5)
            //
            // Another example:
            // s0 = (3,2,1,5) and s1 = (2,7,5)
            // In this case we need to broadcast the (:,:,1:,:) axis
            // of s0 plus the 1st dimension of s1 giving
            // output_lens = (3,2,7,5)
            //
Khalique's avatar
Khalique committed
174
175
176
177
178
179
180
181
            // Get lengths for both arguments
            const std::vector<std::size_t>* s0 = &arg0->get_shape().lens();
            const std::vector<std::size_t>* s1 = &arg1->get_shape().lens();

            // Make sure s0 is the smaller size
            if(s0->size() > s1->size())
                std::swap(s0, s1);

Khalique's avatar
Khalique committed
182
            std::vector<std::size_t> output_lens(*s1);
Khalique's avatar
Khalique committed
183
184
            auto offset = s1->size() - s0->size();
            std::transform(s0->begin(),
Khalique's avatar
Khalique committed
185
186
187
188
                           s0->end(),
                           s1->begin() + offset,
                           output_lens.begin() + offset,
                           [](auto a, auto b) { return std::max(a, b); });
Khalique's avatar
Khalique committed
189
190
191
192
193
194
195
196
197

            auto l0 = prog.add_instruction(op::multibroadcast{output_lens}, arg0);
            auto l1 = prog.add_instruction(op::multibroadcast{output_lens}, arg1);
            return prog.add_instruction(x, l0, l1);
        }
        else
        {
            return prog.add_instruction(x, {arg0, arg1});
        }
198
199
    }

Paul's avatar
Paul committed
200
    template <class T>
Paul's avatar
Paul committed
201
202
    void add_generic_op(std::string name, T x)
    {
Paul's avatar
Paul committed
203
        add_op(name, [this, x](attribute_map, std::vector<instruction_ref> args) {
Paul's avatar
Paul committed
204
205
206
207
            return prog.add_instruction(x, args);
        });
    }

Khalique's avatar
Khalique committed
208
    template <class T>
Khalique's avatar
Khalique committed
209
    void add_variadic_op(std::string name, T x)
Khalique's avatar
Khalique committed
210
    {
Paul's avatar
Paul committed
211
        add_op(name, [this, x](attribute_map, std::vector<instruction_ref> args) {
Khalique's avatar
Khalique committed
212
            return std::accumulate(std::next(args.begin()),
Khalique's avatar
Khalique committed
213
214
215
216
217
                                   args.end(),
                                   args.front(),
                                   [this, x](instruction_ref a, instruction_ref b) {
                                       return add_broadcastable_binary_op(a, b, x);
                                   });
Khalique's avatar
Khalique committed
218
        });
Khalique's avatar
Khalique committed
219
220
    }

Paul's avatar
Paul committed
221
    instruction_ref
Paul's avatar
Paul committed
222
    parse_softmax(const std::string&, const attribute_map&, std::vector<instruction_ref> args)
Paul's avatar
Paul committed
223
224
    {
        auto dims = args.front()->get_shape().lens();
Scott Thornton's avatar
Scott Thornton committed
225
226
        auto r =
            prog.add_instruction(op::reshape{{long(dims[0]), long(dims[1]), 1, 1}}, args.front());
227
228
        auto s = prog.add_instruction(op::softmax{}, r);
        return prog.add_instruction(op::reshape{{long(dims[0]), long(dims[1])}}, s);
Paul's avatar
Paul committed
229
230
    }

Paul's avatar
Paul committed
231
    instruction_ref
Paul's avatar
Paul committed
232
    parse_conv(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Paul's avatar
Paul committed
233
    {
234
        op::convolution op;
235
        auto l0 = args[0];
Paul's avatar
Paul committed
236
237
        if(contains(attributes, "pads"))
        {
Scott Thornton's avatar
Scott Thornton committed
238
            if(contains(attributes, "auto_pad"))
239
            {
Paul's avatar
Paul committed
240
                MIGRAPHX_THROW("auto_pad and padding cannot be specified simultaneously");
241
            }
242
243
            std::vector<std::int64_t> padding;
            copy(attributes["pads"].ints(), std::back_inserter(padding));
Scott Thornton's avatar
Scott Thornton committed
244
            if(padding.size() != 4)
245
            {
Paul's avatar
Paul committed
246
                MIGRAPHX_THROW("padding should have 4 values");
247
            }
Scott Thornton's avatar
Scott Thornton committed
248
            if(padding[0] != padding[2] || padding[1] != padding[3])
249
            {
250
251
                // insert zeros for pad op (args[0] has 4 dims)
                padding = {0, 0, padding[0], padding[1], 0, 0, padding[2], padding[3]};
Khalique's avatar
Khalique committed
252
                l0      = prog.add_instruction(op::pad{padding}, l0);
253
            }
254
255
256
257
            else
            {
                op.padding[0] = padding[0];
                op.padding[1] = padding[1];
258
            }
Paul's avatar
Paul committed
259
        }
Paul's avatar
Paul committed
260
261
262
263
264
265
266
267
        if(contains(attributes, "strides"))
        {
            copy(attributes["strides"].ints(), op.stride.begin());
        }
        if(contains(attributes, "dilations"))
        {
            copy(attributes["dilations"].ints(), op.dilation.begin());
        }
Scott Thornton's avatar
Scott Thornton committed
268
        if(contains(attributes, "auto_pad"))
269
270
        {
            auto s = attributes["auto_pad"].s();
Scott Thornton's avatar
Scott Thornton committed
271
            if(contains(attributes, "pads") and to_upper(s) != "NOTSET")
272
            {
Paul's avatar
Paul committed
273
                MIGRAPHX_THROW("auto_pad and padding cannot be specified simultaneously");
274
275
            }

wsttiger's avatar
fixes  
wsttiger committed
276
            if(s.find("SAME") != std::string::npos)
277
            {
278
                op.padding_mode = op::padding_mode_t::same;
279
280
            }
        }
Khalique's avatar
Khalique committed
281
282
283
284
        if(contains(attributes, "group"))
        {
            op.group = parse_value(attributes.at("group")).at<int>();
        }
Paul's avatar
Paul committed
285
286
287
288
        if(args.size() == 3)
        {
            uint64_t axis = 1;
            auto l1       = prog.add_instruction(op, args[0], args[1]);
Scott Thornton's avatar
Scott Thornton committed
289
            auto l2       = prog.add_instruction(op::broadcast{axis, l1->get_shape()}, args[2]);
290
            return prog.add_instruction(op::add{}, l1, l2);
Paul's avatar
Paul committed
291
        }
292
        return prog.add_instruction(op, l0, args[1]);
Paul's avatar
Paul committed
293
    }
Paul's avatar
Paul committed
294

Paul's avatar
Paul committed
295
296
297
    instruction_ref parse_pooling(const std::string& name,
                                  attribute_map attributes,
                                  std::vector<instruction_ref> args)
Paul's avatar
Paul committed
298
    {
Khalique's avatar
Khalique committed
299
        op::pooling op{ends_with(name, "MaxPool") ? "max" : "average"};
300
        auto l0 = args[0];
Khalique's avatar
Khalique committed
301
        if(starts_with(name, "Global"))
302
        {
Khalique's avatar
Khalique committed
303
304
            auto lens  = args.front()->get_shape().lens();
            op.lengths = {lens[2], lens[3]};
305
        }
Paul's avatar
Paul committed
306
307
        if(contains(attributes, "pads"))
        {
308
309
            std::vector<std::int64_t> padding;
            copy(attributes["pads"].ints(), std::back_inserter(padding));
Scott Thornton's avatar
Scott Thornton committed
310
            if(padding.size() != 4)
311
            {
Paul's avatar
Paul committed
312
                MIGRAPHX_THROW("padding should have 4 values");
313
            }
Scott Thornton's avatar
Scott Thornton committed
314
            if(padding[0] != padding[2] || padding[1] != padding[3])
315
            {
316
317
                // insert zeros for pad op (args[0] has 4 dims)
                padding = {0, 0, padding[0], padding[1], 0, 0, padding[2], padding[3]};
Khalique's avatar
Khalique committed
318
                l0      = prog.add_instruction(op::pad{padding}, l0);
319
320
321
322
323
            }
            else
            {
                op.padding[0] = padding[0];
                op.padding[1] = padding[1];
324
            }
Paul's avatar
Paul committed
325
326
327
328
329
330
331
332
333
        }
        if(contains(attributes, "strides"))
        {
            copy(attributes["strides"].ints(), op.stride.begin());
        }
        if(contains(attributes, "kernel_shape"))
        {
            copy(attributes["kernel_shape"].ints(), op.lengths.begin());
        }
Scott Thornton's avatar
Scott Thornton committed
334
        if(contains(attributes, "auto_pad"))
335
336
        {
            auto s = attributes["auto_pad"].s();
337
            if(s.find("SAME_UPPER") == std::string::npos)
338
            {
339
                MIGRAPHX_THROW("auto_pad only supports SAME_UPPER for pooling");
340
            }
341
            op.padding_mode = op::padding_mode_t::same;
342
343
        }

344
        return prog.add_instruction(op, l0);
Paul's avatar
Paul committed
345
346
    }

Paul's avatar
Paul committed
347
    instruction_ref
Paul's avatar
Paul committed
348
    parse_reshape(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Paul's avatar
Paul committed
349
    {
350
        op::reshape op;
Paul's avatar
Paul committed
351
352
353
354
355
356
357
        if(args.size() == 1)
        {
            literal s = parse_value(attributes.at("shape"));
            s.visit([&](auto v) { copy(v, std::back_inserter(op.dims)); });
        }
        if(args.size() == 2)
        {
Paul's avatar
Paul committed
358
            literal s = args[1]->get_literal();
Paul's avatar
Paul committed
359
            s.visit([&](auto v) { copy(v, std::back_inserter(op.dims)); });
Paul's avatar
Paul committed
360
        }
Paul's avatar
Paul committed
361
362
363
        return prog.add_instruction(op, args[0]);
    }

Paul's avatar
Paul committed
364
    instruction_ref
Paul's avatar
Paul committed
365
    parse_flatten(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Paul's avatar
Paul committed
366
    {
367
        uint64_t axis = 1;
Paul's avatar
Paul committed
368
369
370
371
        if(contains(attributes, "axis"))
        {
            axis = parse_value(attributes.at("axis")).at<int>();
        }
372
        return prog.add_instruction(op::flatten{axis}, args[0]);
Paul's avatar
Paul committed
373
374
    }

375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
    instruction_ref
    parse_squeeze(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        op::squeeze op;
        literal s = parse_value(attributes.at("axes"));
        s.visit([&](auto v) { copy(v, std::back_inserter(op.axes)); });
        return prog.add_instruction(op, args[0]);
    }

    instruction_ref
    parse_unsqueeze(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        op::unsqueeze op;
        literal s = parse_value(attributes.at("axes"));
        s.visit([&](auto v) { copy(v, std::back_inserter(op.axes)); });
        return prog.add_instruction(op, args[0]);
    }

Scott Thornton's avatar
Scott Thornton committed
393
394
395
396
397
398
399
    instruction_ref
    parse_concat(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        std::size_t axis = parse_value(attributes.at("axis")).at<int>();
        op::concat op{axis};
        return prog.add_instruction(op, std::move(args));
    }
400

401
402
403
    instruction_ref
    parse_gather(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
404
        int axis = 0;
405
406
407
408
        if(contains(attributes, "axis"))
        {
            axis = parse_value(attributes.at("axis")).at<int>();
        }
409
        op::gather op{axis};
410
411
412
        return prog.add_instruction(op, std::move(args));
    }

413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
    instruction_ref
    parse_slice(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        op::slice op;
        if(contains(attributes, "axes"))
        {
            literal s = parse_value(attributes.at("axes"));
            s.visit([&](auto v) { copy(v, std::back_inserter(op.axes)); });
        }
        {
            literal s = parse_value(attributes.at("ends"));
            s.visit([&](auto v) { copy(v, std::back_inserter(op.ends)); });
        }
        {
            literal s = parse_value(attributes.at("starts"));
            s.visit([&](auto v) { copy(v, std::back_inserter(op.starts)); });
        }
        return prog.add_instruction(op, args[0]);
    }

Paul's avatar
Paul committed
433
434
435
    instruction_ref parse_constant(const std::string&,
                                   attribute_map attributes,
                                   const std::vector<instruction_ref>&)
Paul's avatar
Paul committed
436
    {
Shucai Xiao's avatar
Shucai Xiao committed
437
        literal v     = parse_value(attributes.at("value"));
438
439
440
        auto dim_size = attributes.at("value").t().dims_size();
        // if dim_size is 0, it is a scalar
        if(dim_size == 0)
441
        {
442
            migraphx::shape scalar_shape{v.get_shape().type(), {1}, {0}};
443
444
445
            return prog.add_literal(migraphx::literal{scalar_shape, v.data()});
        }

Paul's avatar
Paul committed
446
447
        return prog.add_literal(v);
    }
Paul's avatar
Paul committed
448

Paul's avatar
Paul committed
449
    instruction_ref
Paul's avatar
Paul committed
450
    parse_gemm(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Paul's avatar
Paul committed
451
452
    {
        float alpha = 1.0f;
Khalique's avatar
Khalique committed
453
        float beta  = 1.0f;
Paul's avatar
Paul committed
454
455
456
457
458
459
460
461
        bool transa = false;
        bool transb = false;
        if(contains(attributes, "alpha"))
        {
            alpha = parse_value(attributes.at("alpha")).at<float>();
        }
        if(contains(attributes, "beta"))
        {
462
            beta = parse_value(attributes.at("beta")).at<float>();
Paul's avatar
Paul committed
463
464
465
466
467
468
469
470
471
472
        }
        if(contains(attributes, "transA"))
        {
            transa = parse_value(attributes.at("transA")).at<bool>();
        }
        if(contains(attributes, "transB"))
        {
            transb = parse_value(attributes.at("transB")).at<bool>();
        }
        std::vector<int64_t> perm = {1, 0};
473
474
        auto l1 = (transa) ? prog.add_instruction(op::transpose{perm}, args[0]) : args[0];
        auto l2 = (transb) ? prog.add_instruction(op::transpose{perm}, args[1]) : args[1];
Paul's avatar
Paul committed
475
476
        if(args.size() == 3)
        {
Khalique's avatar
Khalique committed
477
            if(beta != 0.f)
478
            {
Khalique's avatar
Khalique committed
479
                auto l3 = prog.add_instruction(op::dot{alpha}, l1, l2);
Khalique's avatar
Khalique committed
480
                auto l4 = args[2];
Khalique's avatar
Khalique committed
481
                if(l4->get_shape().scalar()) // ignore args[2] (no C value added to alpha*A*B)
Khalique's avatar
Khalique committed
482
                    return l3;
Khalique's avatar
Khalique committed
483
                if(beta != 1.f)
Khalique's avatar
Khalique committed
484
485
                {
                    auto beta_val = prog.add_literal(beta);
Khalique's avatar
Khalique committed
486
487
                    auto l5 = prog.add_instruction(op::scalar{args[2]->get_shape()}, beta_val);
                    l4      = prog.add_instruction(op::mul{}, args[2], l5);
Khalique's avatar
Khalique committed
488
489
                }
                return add_broadcastable_binary_op(l3, l4, op::add{});
490
            }
Paul's avatar
Paul committed
491
        }
492
        return prog.add_instruction(op::dot{alpha, beta}, l1, l2);
Paul's avatar
Paul committed
493
494
    }

495
    instruction_ref
Paul's avatar
Paul committed
496
    parse_batchnorm(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
497
    {
Scott Thornton's avatar
Scott Thornton committed
498
499
        float epsilon                                     = 1e-5f;
        float momentum                                    = 0.9f;
500
        op::batch_norm_inference::bn_infer_mode_t bn_mode = op::batch_norm_inference::spatial;
Scott Thornton's avatar
Scott Thornton committed
501
        bool is_test                                      = false;
502
503
504
505
506
507
        if(contains(attributes, "epsilon"))
        {
            epsilon = parse_value(attributes.at("epsilon")).at<float>();
        }
        if(contains(attributes, "momentum"))
        {
508
            momentum = parse_value(attributes.at("momentum")).at<float>();
509
510
511
        }
        if(contains(attributes, "is_test"))
        {
wsttiger's avatar
wsttiger committed
512
            is_test = parse_value(attributes.at("is_test")).at<uint64_t>() > 0;
513
514
515
        }
        if(contains(attributes, "spatial"))
        {
516
            bn_mode = (parse_value(attributes.at("spatial")).at<uint64_t>() > 0)
517
518
                          ? op::batch_norm_inference::spatial
                          : op::batch_norm_inference::per_activation;
519
        }
Paul's avatar
Paul committed
520
        (void)is_test;
Paul's avatar
Paul committed
521
        op::batch_norm_inference op{epsilon, momentum, bn_mode};
Paul's avatar
Paul committed
522
        return prog.add_instruction(op, std::move(args));
523
524
    }

525
526
527
528
    instruction_ref parse_leaky_relu(const std::string&,
                                     attribute_map attributes,
                                     std::vector<instruction_ref> args)
    {
Khalique's avatar
Khalique committed
529
        float alpha = 0.01; // default alpha val for leaky relu
530
531
532
533
534
535
536
537
        if(contains(attributes, "alpha"))
        {
            alpha = parse_value(attributes.at("alpha")).at<float>();
        }
        op::leaky_relu op{alpha};
        return prog.add_instruction(op, args.front());
    }

Khalique's avatar
Khalique committed
538
539
    instruction_ref
    parse_elu(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Khalique's avatar
Khalique committed
540
541
542
543
544
545
546
547
548
549
    {
        float alpha = 1.0; // default alpha val for elu
        if(contains(attributes, "alpha"))
        {
            alpha = parse_value(attributes.at("alpha")).at<float>();
        }
        op::elu op{alpha};
        return prog.add_instruction(op, args.front());
    }

Khalique's avatar
Khalique committed
550
551
    instruction_ref
    parse_lrn(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Khalique's avatar
Khalique committed
552
553
    {
        float alpha = 0.0001;
Khalique's avatar
Khalique committed
554
555
556
        float beta  = 0.75;
        float bias  = 1.0;
        int size    = 1;
Khalique's avatar
Khalique committed
557
558
559
560
561
562
563
564
565
566
567
568
        if(contains(attributes, "alpha"))
            alpha = parse_value(attributes.at("alpha")).at<float>();
        if(contains(attributes, "beta"))
            beta = parse_value(attributes.at("beta")).at<float>();
        if(contains(attributes, "bias"))
            bias = parse_value(attributes.at("bias")).at<float>();
        if(contains(attributes, "size"))
            size = parse_value(attributes.at("size")).at<int>();
        op::lrn op{alpha, beta, bias, size};
        return prog.add_instruction(op, args.front());
    }

Khalique's avatar
Khalique committed
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
    instruction_ref parse_imagescaler(const std::string&,
                                      attribute_map attributes,
                                      std::vector<instruction_ref> args)
    {
        float scale = 1.0;
        std::vector<float> bias{};
        if(contains(attributes, "scale"))
        {
            scale = parse_value(attributes.at("scale")).at<float>();
        }

        if(contains(attributes, "bias"))
        {
            auto&& bias_floats = attributes["bias"].floats();
            bias               = std::vector<float>(bias_floats.begin(), bias_floats.end());
        }
        auto input_shape = args.front()->get_shape();
Khalique's avatar
Khalique committed
586

Khalique's avatar
Khalique committed
587
588
        auto scale_val = prog.add_literal(scale);
        auto bias_vals = prog.add_literal(
Paul's avatar
Paul committed
589
            migraphx::literal{migraphx::shape{migraphx::shape::float_type, {bias.size()}}, bias});
Khalique's avatar
Khalique committed
590

Paul's avatar
Paul committed
591
592
        auto scale_tensor = prog.add_instruction(migraphx::op::scalar{input_shape}, scale_val);
        auto img_scaled   = prog.add_instruction(migraphx::op::mul{}, args.front(), scale_tensor);
Paul's avatar
Paul committed
593
        auto bias_bcast = prog.add_instruction(migraphx::op::broadcast{1, input_shape}, bias_vals);
Paul's avatar
Paul committed
594
        return prog.add_instruction(migraphx::op::add{}, img_scaled, bias_bcast);
Khalique's avatar
Khalique committed
595
    }
Khalique's avatar
Khalique committed
596

Khalique's avatar
Khalique committed
597
598
    instruction_ref
    parse_transpose(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
Khalique's avatar
Khalique committed
599
600
601
602
603
604
605
    {
        std::vector<int64_t> perm{};
        if(contains(attributes, "perm"))
        {
            auto&& perm_vals = attributes["perm"].ints();
            perm             = std::vector<int64_t>(perm_vals.begin(), perm_vals.end());
        }
Paul's avatar
Paul committed
606
        return prog.add_instruction(migraphx::op::transpose{perm}, args.front());
Khalique's avatar
Khalique committed
607
608
    }

Khalique's avatar
Khalique committed
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
    instruction_ref
    parse_pad(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        std::vector<int64_t> pads{};
        float value = 0.0f;
        if(contains(attributes, "pads"))
        {
            auto&& pad_vals = attributes["pads"].ints();
            pads            = std::vector<int64_t>(pad_vals.begin(), pad_vals.end());
        }
        if(contains(attributes, "value"))
        {
            value = parse_value(attributes.at("value")).at<float>();
        }
        if(contains(attributes, "mode"))
        {
            auto mode = attributes.at("mode").s();
            if(mode != "constant")
                MIGRAPHX_THROW("migraphx currently only supports constant padding");
        }
        return prog.add_instruction(migraphx::op::pad{pads, value}, args.front());
    }
631
632
633
    // Use a literal instruction to replace the shape since, output of
    // shape operator are literals in migraphx
    instruction_ref
Shucai Xiao's avatar
Shucai Xiao committed
634
    parse_shape(const std::string&, const attribute_map&, std::vector<instruction_ref> args)
635
636
    {
        if(args.size() != 1)
637
            MIGRAPHX_THROW("Shape: operator should have 1 operand");
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
        std::vector<std::size_t> arg_shape = args[0]->get_shape().lens();
        std::vector<int64_t> vec_shape(arg_shape.size());
        migraphx::shape s(migraphx::shape::int64_type, {arg_shape.size()});
        std::transform(arg_shape.begin(), arg_shape.end(), vec_shape.begin(), [](auto i) {
            return int64_t(i);
        });
        return prog.add_literal(migraphx::literal{s, vec_shape});
    }

    // Use a literal instruction to replace the constantFill operator. In RNN, input shape
    // and value are fixed, so no need to do the actual computation for the constantFill
    // operator
    instruction_ref parse_constant_fill(const std::string&,
                                        attribute_map attributes,
                                        std::vector<instruction_ref> args)
    {
        int input_as_shape = 0;
        int dtype          = 1;
        float value        = 0.0f;

        if(contains(attributes, "dtype"))
        {
            dtype = parse_value(attributes.at("dtype")).at<int>();
        }
        migraphx::shape::type_t type = get_type(dtype);

        if(contains(attributes, "input_as_shape"))
        {
            input_as_shape = parse_value(attributes.at("input_as_shape")).at<int>();
        }

        if(contains(attributes, "value"))
        {
            value = parse_value(attributes.at("value")).at<float>();
        }

Shucai Xiao's avatar
Shucai Xiao committed
674
675
        if(contains(attributes, "extra_shape"))
        {
676
            MIGRAPHX_THROW("ConstantFill: cannot handle extra shape attribute");
677
678
        }

679
680
        if(input_as_shape == 1)
        {
Shucai Xiao's avatar
Shucai Xiao committed
681
            if(args.size() != 1)
682
            {
683
                MIGRAPHX_THROW("ConstantFill: need an input argument as output shape");
684
685
            }

Shucai Xiao's avatar
Shucai Xiao committed
686
687
            if(contains(attributes, "shape"))
            {
688
                MIGRAPHX_THROW("ConstantFill: cannot set the shape argument and pass in an input "
Shucai Xiao's avatar
Shucai Xiao committed
689
                               "at the same time");
690
691
            }

692
693
694
            migraphx::argument in = args[0]->eval();
            if(in.empty())
            {
695
                MIGRAPHX_THROW("ConstantFill: cannot handle dynamic shape as input");
696
            }
697

698
699
700
            std::vector<std::size_t> dims;
            in.visit([&](auto input) { dims.assign(input.begin(), input.end()); });
            migraphx::shape s(type, dims);
701
702
            std::vector<float> values(s.elements(), value);
            return prog.add_literal(migraphx::literal(s, values));
703
704
705
        }
        else if(input_as_shape == 0)
        {
Shucai Xiao's avatar
Shucai Xiao committed
706
707
            if(!contains(attributes, "shape"))
            {
708
                MIGRAPHX_THROW("ConstantFill: attribute output shape is needed");
709
710
711
            }

            literal ls = parse_value(attributes.at("shape"));
712
            std::vector<std::size_t> dims;
Shucai Xiao's avatar
Shucai Xiao committed
713
            ls.visit([&](auto s) { dims.assign(s.begin(), s.end()); });
714
            migraphx::shape s{type, dims};
715
716
            std::vector<float> values(s.elements(), value);
            return prog.add_literal(migraphx::literal(s, values));
717
718
719
        }
        else
        {
720
            MIGRAPHX_THROW("ConstantFill: wrong value of attribute input_as_shape");
721
722
723
        }
    }

Shucai Xiao's avatar
Shucai Xiao committed
724
    std::vector<instruction_ref>
Shucai Xiao's avatar
Shucai Xiao committed
725
726
727
    parse_rnn(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        migraphx::shape input_shape = args[0]->get_shape();
728
        std::size_t hidden_size     = args[1]->get_shape().lens()[1];
Shucai Xiao's avatar
Shucai Xiao committed
729
730
731

        if(contains(attributes, "hidden_size"))
        {
Shucai Xiao's avatar
Shucai Xiao committed
732
            std::size_t hidden_size_att = parse_value(attributes.at("hidden_size")).at<int>();
Shucai Xiao's avatar
Shucai Xiao committed
733
            if(hidden_size != hidden_size_att)
Shucai Xiao's avatar
Shucai Xiao committed
734
735
736
            {
                MIGRAPHX_THROW("RNN: hidden size mismatch in input and attribute");
            }
Shucai Xiao's avatar
Shucai Xiao committed
737
738
739
740
741
742
743
744
745
        }

        // Handling of direction to be added later
        std::string direction{"forward"};
        if(contains(attributes, "direction"))
        {
            direction = attributes.at("direction").s();
        }

746
        op::rnn_direction dirct = op::rnn_direction::forward;
Shucai Xiao's avatar
Shucai Xiao committed
747
748
        if(direction == "bidirectional")
        {
749
            dirct = op::rnn_direction::bidirectional;
Shucai Xiao's avatar
Shucai Xiao committed
750
751
752
        }
        else if(direction == "reverse")
        {
753
            dirct = op::rnn_direction::reverse;
Shucai Xiao's avatar
Shucai Xiao committed
754
755
        }

756
757
758
759
760
        std::vector<std::string> vec_names{"tanh"};
        if(contains(attributes, "activations"))
        {
            auto names = attributes.at("activations").strings();
            vec_names.clear();
761
            vec_names.resize(names.size());
762
            std::copy(names.begin(), names.end(), vec_names.begin());
763
764
        }

765
766
767
        auto name_it = std::find_if(vec_names.begin(), vec_names.end(), [&](auto& name) {
            return (map_actv_funcs.count(name) == 0);
        });
Shucai Xiao's avatar
Shucai Xiao committed
768
        if(name_it != vec_names.end())
769
770
771
        {
            MIGRAPHX_THROW("RNN: activation function " + std::string(*name_it) + " not supported");
        }
772

Shucai Xiao's avatar
Shucai Xiao committed
773
        // bidirectional case should have two activation functions.
Shucai Xiao's avatar
Shucai Xiao committed
774
        // one is for forward, and the other is for reverse.
Shucai Xiao's avatar
Shucai Xiao committed
775
        // if only one actv function is provided, we use it in both
776
        // forward and reverse direction
777
        if(dirct == op::rnn_direction::bidirectional)
778
        {
Shucai Xiao's avatar
Shucai Xiao committed
779
            if(vec_names.size() == 1)
780
781
782
783
784
            {
                vec_names.push_back(vec_names.at(0));
            }
        }

Shucai Xiao's avatar
Shucai Xiao committed
785
786
787
        std::vector<operation> vec_actv_funcs(vec_names.size());
        std::transform(vec_names.begin(), vec_names.end(), vec_actv_funcs.begin(), [&](auto& fn) {
            return map_actv_funcs[fn];
788
        });
Shucai Xiao's avatar
Shucai Xiao committed
789

Shucai Xiao's avatar
Shucai Xiao committed
790
791
792
793
794
795
796
        // To be added later
        float clip = 0.0;
        if(contains(attributes, "clip"))
        {
            clip = parse_value(attributes.at("clip")).at<float>();
        }

797
798
        // if the number of arguments is less than 6, append
        // undefined operator to have 6 arguments
Shucai Xiao's avatar
Shucai Xiao committed
799
        if(args.size() < 6)
800
801
802
803
804
        {
            auto ins = prog.add_instruction(op::undefined{});
            args.insert(args.end(), (6 - args.size()), ins);
        }

Shucai Xiao's avatar
Shucai Xiao committed
805
806
        // first output for the concatenation of hidden states
        auto hidden_states = prog.add_instruction(op::rnn{hidden_size, vec_actv_funcs, dirct, clip},
Shucai Xiao's avatar
Shucai Xiao committed
807
                                                  std::move(args));
Shucai Xiao's avatar
Shucai Xiao committed
808

809
        // second output for the last hidden state
Shucai Xiao's avatar
Shucai Xiao committed
810
        auto last_output = prog.add_instruction(op::rnn_last_output{}, hidden_states);
Shucai Xiao's avatar
Shucai Xiao committed
811

Shucai Xiao's avatar
Shucai Xiao committed
812
        return {hidden_states, last_output};
Shucai Xiao's avatar
Shucai Xiao committed
813
814
    }

815
    std::vector<instruction_ref>
816
817
818
819
820
821
822
    parse_gru(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        migraphx::shape input_shape = args[0]->get_shape();
        std::size_t hidden_size     = args[2]->get_shape().lens()[2];

        if(contains(attributes, "hidden_size"))
        {
Shucai Xiao's avatar
Shucai Xiao committed
823
            std::size_t hidden_size_att = parse_value(attributes.at("hidden_size")).at<int>();
Shucai Xiao's avatar
Shucai Xiao committed
824
            if(hidden_size != hidden_size_att)
Shucai Xiao's avatar
Shucai Xiao committed
825
826
827
            {
                MIGRAPHX_THROW("GRU: hidden size mismatch in input and attribute");
            }
828
829
830
831
832
833
834
835
836
        }

        // Handling of direction to be added later
        std::string direction{"forward"};
        if(contains(attributes, "direction"))
        {
            direction = attributes.at("direction").s();
        }

837
        op::rnn_direction dirct = op::rnn_direction::forward;
838
839
        if(direction == "bidirectional")
        {
840
            dirct = op::rnn_direction::bidirectional;
841
842
843
        }
        else if(direction == "reverse")
        {
844
            dirct = op::rnn_direction::reverse;
845
846
        }

847
        std::vector<std::string> vec_names = {"sigmoid", "tanh"};
848
849
        if(contains(attributes, "activations"))
        {
850
            auto names = attributes.at("activations").strings();
851
            vec_names.clear();
Shucai Xiao's avatar
Shucai Xiao committed
852
            vec_names.resize(names.size());
853
            std::copy(names.begin(), names.end(), vec_names.begin());
854
855
        }

856
        // need 4 activation functions
857
        if(dirct == op::rnn_direction::bidirectional)
858
        {
Shucai Xiao's avatar
Shucai Xiao committed
859
            // 4 activation functions are used in the bidirectional
860
            // scenario. No spec is provided in onnx::operator. we
Shucai Xiao's avatar
Shucai Xiao committed
861
862
            // use the algorithm that: if 1 actv function is provided,
            // repeat 1 four times. If 2 actv functins are provided,
863
864
            // assume forward and reverse use the same pair of actv
            // functions. For the case of 3 actv functions provided,
Shucai Xiao's avatar
Shucai Xiao committed
865
866
867
            // assume the 3rd one is repeated once and used by the
            // reverse direction.
            // This may need change later
868
            if(vec_names.size() == 1)
869
            {
870
                vec_names.insert(vec_names.end(), 3, vec_names.at(0));
871
            }
872
            else if(vec_names.size() == 2)
873
            {
874
875
876
                // repeat the activation functions
                vec_names.push_back(vec_names.at(0));
                vec_names.push_back(vec_names.at(1));
877
            }
878
            else if(vec_names.size() == 3)
879
            {
880
                vec_names.push_back(vec_names.at(2));
881
882
            }
        }
Shucai Xiao's avatar
Shucai Xiao committed
883
        else
884
        {
885
            if(vec_names.size() == 1)
886
            {
887
                vec_names.push_back(vec_names.at(0));
888
889
890
            }
        }

891
892
893
        auto name_it = std::find_if(vec_names.begin(), vec_names.end(), [&](auto& name) {
            return (map_actv_funcs.count(name) == 0);
        });
Shucai Xiao's avatar
Shucai Xiao committed
894
        if(name_it != vec_names.end())
895
896
897
        {
            MIGRAPHX_THROW("GRU: activation function " + std::string(*name_it) + " not supported");
        }
898

Shucai Xiao's avatar
Shucai Xiao committed
899
900
901
        std::vector<operation> vec_actv_funcs(vec_names.size());
        std::transform(vec_names.begin(), vec_names.end(), vec_actv_funcs.begin(), [&](auto& name) {
            return map_actv_funcs[name];
Shucai Xiao's avatar
Shucai Xiao committed
902
        });
903
904
905
906
907
908
909
910

        float clip = 0.0;
        if(contains(attributes, "clip"))
        {
            clip = parse_value(attributes.at("clip")).at<float>();
        }

        int linear_before_reset = 0;
Shucai Xiao's avatar
Shucai Xiao committed
911
        if(contains(attributes, "linear_before_reset"))
912
913
914
915
        {
            linear_before_reset = parse_value(attributes.at("linear_before_reset")).at<int>();
        }

Shucai Xiao's avatar
Shucai Xiao committed
916
        // append undefined opeator to make 6 arguments
Shucai Xiao's avatar
Shucai Xiao committed
917
        if(args.size() < 6)
Shucai Xiao's avatar
Shucai Xiao committed
918
919
920
921
922
        {
            auto ins = prog.add_instruction(op::undefined{});
            args.insert(args.end(), 6 - args.size(), ins);
        }

923
924
        // first output for concatenation of hidden states
        auto hidden_states = prog.add_instruction(
Shucai Xiao's avatar
Shucai Xiao committed
925
            op::gru{hidden_size, vec_actv_funcs, dirct, clip, linear_before_reset},
Shucai Xiao's avatar
Shucai Xiao committed
926
            std::move(args));
927
928

        // second output for last gru output
929
        auto last_output = prog.add_instruction(op::rnn_last_output{}, hidden_states);
930

Shucai Xiao's avatar
Shucai Xiao committed
931
        return {hidden_states, last_output};
932
933
    }

Shucai Xiao's avatar
Shucai Xiao committed
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
    std::vector<instruction_ref>
    parse_lstm(const std::string&, attribute_map attributes, std::vector<instruction_ref> args)
    {
        migraphx::shape input_shape = args[0]->get_shape();
        std::size_t hidden_size     = args[2]->get_shape().lens()[2];

        if(contains(attributes, "hidden_size"))
        {
            std::size_t hidden_size_att = parse_value(attributes.at("hidden_size")).at<int>();
            if(hidden_size != hidden_size_att)
            {
                MIGRAPHX_THROW("LSTM: hidden size mismatch in input and attribute");
            }
        }

        // Handling of direction to be added later
        std::string direction{"forward"};
        if(contains(attributes, "direction"))
        {
            direction = attributes.at("direction").s();
        }

Shucai Xiao's avatar
Shucai Xiao committed
956
        op::rnn_direction dirct = op::rnn_direction::forward;
Shucai Xiao's avatar
Shucai Xiao committed
957
958
        if(direction == "bidirectional")
        {
Shucai Xiao's avatar
Shucai Xiao committed
959
            dirct = op::rnn_direction::bidirectional;
Shucai Xiao's avatar
Shucai Xiao committed
960
961
962
        }
        else if(direction == "reverse")
        {
Shucai Xiao's avatar
Shucai Xiao committed
963
            dirct = op::rnn_direction::reverse;
Shucai Xiao's avatar
Shucai Xiao committed
964
        }
Shucai Xiao's avatar
Shucai Xiao committed
965
        else if(direction == "forward")
Shucai Xiao's avatar
Shucai Xiao committed
966
        {
Shucai Xiao's avatar
Shucai Xiao committed
967
            dirct = op::rnn_direction::forward;
Shucai Xiao's avatar
Shucai Xiao committed
968
969
970
971
972
973
974
975
976
977
978
979
        }
        else
        {
            MIGRAPHX_THROW("LSTM: incorrect direction attribute");
        }

        std::vector<std::string> vec_names = {"sigmoid", "tanh", "tanh"};
        if(contains(attributes, "activations"))
        {
            auto names = attributes.at("activations").strings();
            vec_names.clear();
            vec_names.resize(names.size());
980
            std::copy(names.begin(), names.end(), vec_names.begin());
Shucai Xiao's avatar
Shucai Xiao committed
981
982
983
        }

        // need 6 activation functions for bidirectional directions
Shucai Xiao's avatar
Shucai Xiao committed
984
        if(dirct == op::rnn_direction::bidirectional)
Shucai Xiao's avatar
Shucai Xiao committed
985
986
987
988
989
990
        {
            // 6 activation functions are used in the bidirectional
            // scenario. No spec is provided in onnx::operator. we
            // use the algorithm that: if 1 actv function is provided,
            // repeat 1st six times. If 2 actv functins are provided,
            // repeat 2nd once, then repeat all three once
Shucai Xiao's avatar
Shucai Xiao committed
991
            // if 3 actv funcs are provide, repeat all three once.
Shucai Xiao's avatar
Shucai Xiao committed
992
993
994
995
            // the same algorithm is used for 4, 5, and 6 actv funcions
            // provided. This may need change later
            switch(vec_names.size())
            {
996
            case 1:
Shucai Xiao's avatar
Shucai Xiao committed
997
998
999
1000
1001
1002
                vec_names = {vec_names.at(0),
                             vec_names.at(0),
                             vec_names.at(0),
                             vec_names.at(0),
                             vec_names.at(0),
                             vec_names.at(0)};
1003
                break;
Shucai Xiao's avatar
Shucai Xiao committed
1004
1005
1006

            case 2:
                // repeat the 2nd actv func once, then repeat all three another time
Shucai Xiao's avatar
Shucai Xiao committed
1007
1008
1009
1010
1011
1012
                vec_names = {vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(1),
                             vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(1)};
Shucai Xiao's avatar
Shucai Xiao committed
1013
1014
1015
1016
                break;

            case 3:
                // repeat all three actv funcs once
Shucai Xiao's avatar
Shucai Xiao committed
1017
1018
1019
1020
1021
1022
                vec_names = {vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(2),
                             vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(2)};
Shucai Xiao's avatar
Shucai Xiao committed
1023
1024
                break;

Shucai Xiao's avatar
Shucai Xiao committed
1025
1026
1027
1028
1029
1030
1031
            case 4:
                vec_names = {vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(2),
                             vec_names.at(3),
                             vec_names.at(3),
                             vec_names.at(3)};
1032
                break;
Shucai Xiao's avatar
Shucai Xiao committed
1033

Shucai Xiao's avatar
Shucai Xiao committed
1034
1035
1036
1037
1038
1039
1040
            case 5:
                vec_names = {vec_names.at(0),
                             vec_names.at(1),
                             vec_names.at(2),
                             vec_names.at(3),
                             vec_names.at(4),
                             vec_names.at(4)};
1041
                break;
Shucai Xiao's avatar
Shucai Xiao committed
1042

Shucai Xiao's avatar
Shucai Xiao committed
1043
            default: break;
Shucai Xiao's avatar
Shucai Xiao committed
1044
1045
1046
1047
1048
1049
            }
        }
        else
        {
            switch(vec_names.size())
            {
Shucai Xiao's avatar
Shucai Xiao committed
1050
            case 1: vec_names = {vec_names.at(0), vec_names.at(0), vec_names.at(0)}; break;
Shucai Xiao's avatar
Shucai Xiao committed
1051
1052
1053

            case 2:
                // repeat the 2nd actv func once, so we have 3 actv funcs
Shucai Xiao's avatar
Shucai Xiao committed
1054
                vec_names = {vec_names.at(0), vec_names.at(1), vec_names.at(1)};
Shucai Xiao's avatar
Shucai Xiao committed
1055
1056
                break;

Shucai Xiao's avatar
Shucai Xiao committed
1057
            default: break;
Shucai Xiao's avatar
Shucai Xiao committed
1058
1059
1060
            }
        }

1061
1062
1063
        auto name_it = std::find_if(vec_names.begin(), vec_names.end(), [&](auto& name) {
            return (map_actv_funcs.count(name) == 0);
        });
Shucai Xiao's avatar
Shucai Xiao committed
1064
        if(name_it != vec_names.end())
1065
1066
1067
        {
            MIGRAPHX_THROW("LSTM: activation function " + std::string(*name_it) + " not supported");
        }
Shucai Xiao's avatar
Shucai Xiao committed
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089

        std::vector<operation> vec_actv_funcs(vec_names.size());
        std::transform(vec_names.begin(), vec_names.end(), vec_actv_funcs.begin(), [&](auto& name) {
            return map_actv_funcs[name];
        });

        float clip = 0.0;
        if(contains(attributes, "clip"))
        {
            clip = parse_value(attributes.at("clip")).at<float>();
        }

        int input_forget = 0;
        if(contains(attributes, "input_forget"))
        {
            input_forget = parse_value(attributes.at("input_forget")).at<int>();
        }

        // append undefined opeator to make 6 arguments
        if(args.size() < 8)
        {
            auto ins = prog.add_instruction(op::undefined{});
Shucai Xiao's avatar
Shucai Xiao committed
1090
            args.insert(args.end(), 8 - args.size(), ins);
Shucai Xiao's avatar
Shucai Xiao committed
1091
1092
1093
1094
        }

        // first output for concatenation of hidden states
        auto hidden_states = prog.add_instruction(
Shucai Xiao's avatar
Shucai Xiao committed
1095
            op::lstm{hidden_size, vec_actv_funcs, dirct, clip, input_forget}, std::move(args));
Shucai Xiao's avatar
Shucai Xiao committed
1096
1097

        // second output for last lstm output
Shucai Xiao's avatar
Shucai Xiao committed
1098
        auto last_output = prog.add_instruction(op::rnn_last_output{}, hidden_states);
Shucai Xiao's avatar
Shucai Xiao committed
1099
1100
1101
1102
1103
1104
1105

        // third output for last cell output
        auto last_cell_output = prog.add_instruction(op::lstm_last_cell_output{}, hidden_states);

        return {hidden_states, last_output, last_cell_output};
    }

Paul's avatar
Paul committed
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
    void parse_from(std::istream& is)
    {
        onnx::ModelProto model;
        if(model.ParseFromIstream(&is))
        {
            if(model.has_graph())
            {
                this->parse_graph(model.graph());
            }
        }
        else
        {
Paul's avatar
Paul committed
1118
            MIGRAPHX_THROW("Failed reading onnx file.");
Paul's avatar
Paul committed
1119
1120
1121
1122
1123
1124
        }
    }

    void parse_graph(const onnx::GraphProto& graph)
    {
        nodes = get_nodes(graph);
1125
1126
1127
1128
1129
        std::unordered_map<std::string, onnx::TensorProto> initializer_data;
        for(auto&& f : graph.initializer())
        {
            initializer_data[f.name()] = f;
        }
Paul's avatar
Paul committed
1130
1131
1132
        for(auto&& input : graph.input())
        {
            const std::string& name = input.name();
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
            // Does the input have an initializer?
            if(contains(initializer_data, name))
            {
                auto t             = initializer_data[name];
                instructions[name] = prog.add_literal(parse_tensor(t));
            }
            else
            {
                // TODO: Get shape of input parameter
                shape s            = parse_type(input.type());
                instructions[name] = prog.add_parameter(name, s);
            }
Paul's avatar
Paul committed
1145
1146
1147
        }
        for(auto&& p : nodes)
        {
Paul's avatar
Paul committed
1148
            this->parse_node(p.first);
Paul's avatar
Paul committed
1149
1150
1151
        }
    }

Shucai Xiao's avatar
Shucai Xiao committed
1152
    void parse_undefined(const std::string& name)
1153
    {
Shucai Xiao's avatar
Shucai Xiao committed
1154
        auto ins           = prog.add_instruction(op::undefined{});
1155
1156
1157
        instructions[name] = ins;
    }

Paul's avatar
Paul committed
1158
    void parse_node(const std::string& name)
Paul's avatar
Paul committed
1159
    {
Paul's avatar
Paul committed
1160
        if(name.empty())
Paul's avatar
Paul committed
1161
            MIGRAPHX_THROW("Onnx node must have a name");
Paul's avatar
Paul committed
1162
1163
1164
1165
1166
1167
1168
1169
        if(instructions.count(name) == 0)
        {
            auto&& node = nodes.at(name);
            std::vector<instruction_ref> args;
            for(auto&& input : node.input())
            {
                if(nodes.count(input) > 0)
                {
Paul's avatar
Paul committed
1170
1171
                    assert(name != input);
                    this->parse_node(input);
Paul's avatar
Paul committed
1172
                }
Shucai Xiao's avatar
Shucai Xiao committed
1173
                else if(input.empty())
Paul's avatar
Paul committed
1174
                {
1175
                    this->parse_undefined(input);
Paul's avatar
Paul committed
1176
                }
1177
                args.push_back(instructions.at(input));
Paul's avatar
Paul committed
1178
            }
Paul's avatar
Paul committed
1179
            std::vector<instruction_ref> result;
Paul's avatar
Paul committed
1180
1181
            if(ops.count(node.op_type()) == 0)
            {
Paul's avatar
Paul committed
1182
                result.push_back(prog.add_instruction(unknown{node.op_type()}, args));
Paul's avatar
Paul committed
1183
1184
1185
            }
            else
            {
Paul's avatar
Paul committed
1186
                result = ops[node.op_type()](get_attributes(node), args);
Paul's avatar
Paul committed
1187
            }
Paul's avatar
Paul committed
1188
            // Even no output nodes produce output in migraphx
Paul's avatar
Paul committed
1189
            if(node.output().empty() and result.size() == 1)
Paul's avatar
Paul committed
1190
1191
            {
                instructions[name] = result.front();
Paul's avatar
Paul committed
1192
1193
1194
            }
            else
            {
Paul's avatar
Paul committed
1195
1196
1197
1198
1199
1200
                assert(node.output().size() >= result.size());
                std::transform(result.begin(),
                               result.end(),
                               node.output().begin(),
                               std::inserter(instructions, instructions.end()),
                               [](auto&& x, auto&& y) { return std::make_pair(y, x); });
Paul's avatar
Paul committed
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
            }
        }
    }

    static attribute_map get_attributes(const onnx::NodeProto& node)
    {
        std::unordered_map<std::string, onnx::AttributeProto> result;
        for(auto&& attr : node.attribute())
        {
            result[attr.name()] = attr;
        }
        return result;
    }

    static node_map get_nodes(const onnx::GraphProto& graph)
    {
        std::unordered_map<std::string, onnx::NodeProto> result;
Paul's avatar
Paul committed
1218
        std::size_t n = 0;
Paul's avatar
Paul committed
1219
1220
        for(auto&& node : graph.node())
        {
Paul's avatar
Paul committed
1221
            if(node.output().empty())
Paul's avatar
Paul committed
1222
            {
Paul's avatar
Paul committed
1223
                if(node.name().empty())
Paul's avatar
Paul committed
1224
1225
1226
1227
1228
1229
1230
1231
1232
                {
                    result["migraphx_unamed_node_" + std::to_string(n)] = node;
                    n++;
                }
                else
                {
                    result[node.name()] = node;
                }
            }
Paul's avatar
Paul committed
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
            for(auto&& output : node.output())
            {
                result[output] = node;
            }
        }
        return result;
    }

    template <class T>
    static literal from_repeated(shape::type_t t, const T& r)
    {
        std::size_t size = r.size();
        return literal{{t, {size}}, r.begin(), r.end()};
    }

    static literal parse_value(const onnx::AttributeProto& attr)
    {
        switch(attr.type())
        {
        case onnx::AttributeProto::UNDEFINED: return {};
        case onnx::AttributeProto::FLOAT: return literal{attr.f()};
        case onnx::AttributeProto::INT: return literal{attr.i()};
        case onnx::AttributeProto::STRING: return {};
        case onnx::AttributeProto::TENSOR: return parse_tensor(attr.t());
        case onnx::AttributeProto::GRAPH: return {};
Paul's avatar
Paul committed
1258
        case onnx::AttributeProto::FLOATS: return from_repeated(shape::float_type, attr.floats());
Paul's avatar
Paul committed
1259
1260
1261
1262
1263
        case onnx::AttributeProto::INTS: return from_repeated(shape::int64_type, attr.ints());
        case onnx::AttributeProto::STRINGS: return {};
        case onnx::AttributeProto::TENSORS: return {};
        case onnx::AttributeProto::GRAPHS: return {};
        }
Paul's avatar
Paul committed
1264
        MIGRAPHX_THROW("Invalid attribute type");
Paul's avatar
Paul committed
1265
1266
1267
1268
1269
    }

    static literal parse_tensor(const onnx::TensorProto& t)
    {
        std::vector<std::size_t> dims(t.dims().begin(), t.dims().end());
Khalique's avatar
Khalique committed
1270
        // in case of scalar constants in onnx file, use dims=1 to fill initializer data
1271
        if(dims.empty())
Khalique's avatar
Khalique committed
1272
1273
1274
        {
            dims = {1};
        }
1275
1276
        if(t.has_raw_data())
        {
wsttiger's avatar
wsttiger committed
1277
            const std::string& s = t.raw_data();
Scott Thornton's avatar
Scott Thornton committed
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
            switch(t.data_type())
            {
            case onnx::TensorProto::UNDEFINED: throw std::runtime_error("");
            case onnx::TensorProto::FLOAT: return literal{{shape::float_type, dims}, s.data()};
            case onnx::TensorProto::UINT8: throw std::runtime_error("");
            case onnx::TensorProto::INT8: return literal{{shape::int32_type, dims}, s.data()};
            case onnx::TensorProto::UINT16: return literal{{shape::int32_type, dims}, s.data()};
            case onnx::TensorProto::INT16: return literal{{shape::int32_type, dims}, s.data()};
            case onnx::TensorProto::INT32: return literal{{shape::int32_type, dims}, s.data()};
            case onnx::TensorProto::INT64: return literal{{shape::int64_type, dims}, s.data()};
            case onnx::TensorProto::STRING: throw std::runtime_error("");
            case onnx::TensorProto::BOOL: return literal{{shape::int32_type, dims}, s.data()};
Paul's avatar
Paul committed
1290
            case onnx::TensorProto::FLOAT16: return literal{{shape::half_type, dims}, s.data()};
Scott Thornton's avatar
Scott Thornton committed
1291
1292
1293
1294
1295
1296
            case onnx::TensorProto::DOUBLE: return literal{{shape::double_type, dims}, s.data()};
            case onnx::TensorProto::UINT32: throw std::runtime_error("");
            case onnx::TensorProto::UINT64: throw std::runtime_error("");
            case onnx::TensorProto::COMPLEX64: throw std::runtime_error("");
            case onnx::TensorProto::COMPLEX128: throw std::runtime_error("");
            }
Paul's avatar
Paul committed
1297
            MIGRAPHX_THROW("Invalid tensor type");
1298
        }
Paul's avatar
Paul committed
1299
1300
1301
1302
        switch(t.data_type())
        {
        case onnx::TensorProto::UNDEFINED: throw std::runtime_error("");
        case onnx::TensorProto::FLOAT:
Paul's avatar
Paul committed
1303
            return literal{{shape::float_type, dims}, t.float_data().begin(), t.float_data().end()};
Paul's avatar
Paul committed
1304
1305
        case onnx::TensorProto::UINT8: throw std::runtime_error("");
        case onnx::TensorProto::INT8:
Paul's avatar
Paul committed
1306
            return literal{{shape::int32_type, dims}, t.int32_data().begin(), t.int32_data().end()};
Paul's avatar
Paul committed
1307
        case onnx::TensorProto::UINT16:
Paul's avatar
Paul committed
1308
            return literal{{shape::int32_type, dims}, t.int32_data().begin(), t.int32_data().end()};
Paul's avatar
Paul committed
1309
        case onnx::TensorProto::INT16:
Paul's avatar
Paul committed
1310
            return literal{{shape::int32_type, dims}, t.int32_data().begin(), t.int32_data().end()};
Paul's avatar
Paul committed
1311
        case onnx::TensorProto::INT32:
Paul's avatar
Paul committed
1312
            return literal{{shape::int32_type, dims}, t.int32_data().begin(), t.int32_data().end()};
Paul's avatar
Paul committed
1313
        case onnx::TensorProto::INT64:
Paul's avatar
Paul committed
1314
            return literal{{shape::int64_type, dims}, t.int64_data().begin(), t.int64_data().end()};
Paul's avatar
Paul committed
1315
1316
        case onnx::TensorProto::STRING: throw std::runtime_error("");
        case onnx::TensorProto::BOOL:
Paul's avatar
Paul committed
1317
            return literal{{shape::int32_type, dims}, t.int32_data().begin(), t.int32_data().end()};
Paul's avatar
Paul committed
1318
        case onnx::TensorProto::FLOAT16:
Khalique's avatar
Khalique committed
1319
        {
Khalique's avatar
Khalique committed
1320
            std::vector<uint16_t> data_uint16(t.int32_data().begin(), t.int32_data().end());
1321
            std::vector<half> data_half;
Khalique's avatar
Khalique committed
1322
1323
1324
            std::transform(data_uint16.begin(),
                           data_uint16.end(),
                           std::back_inserter(data_half),
1325
                           [](uint16_t raw_val) { return *reinterpret_cast<half*>(&raw_val); });
1326
            return literal{{shape::half_type, dims}, data_half.begin(), data_half.end()};
Khalique's avatar
Khalique committed
1327
        }
Paul's avatar
Paul committed
1328
1329
1330
1331
1332
1333
1334
1335
        case onnx::TensorProto::DOUBLE:
            return literal{
                {shape::double_type, dims}, t.double_data().begin(), t.double_data().end()};
        case onnx::TensorProto::UINT32: throw std::runtime_error("");
        case onnx::TensorProto::UINT64: throw std::runtime_error("");
        case onnx::TensorProto::COMPLEX64: throw std::runtime_error("");
        case onnx::TensorProto::COMPLEX128: throw std::runtime_error("");
        }
Paul's avatar
Paul committed
1336
        MIGRAPHX_THROW("Invalid tensor type");
Paul's avatar
Paul committed
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
    }

    static shape parse_type(const onnx::TypeProto& t)
    {
        shape::type_t shape_type{};
        switch(t.tensor_type().elem_type())
        {
        case onnx::TensorProto::UNDEFINED:
            break; // throw std::runtime_error("Unsupported type UNDEFINED");
        case onnx::TensorProto::FLOAT: shape_type = shape::float_type; break;
        case onnx::TensorProto::UINT8:
            break; // throw std::runtime_error("Unsupported type UINT8");
        case onnx::TensorProto::INT8: shape_type = shape::int8_type; break;
        case onnx::TensorProto::UINT16: shape_type = shape::uint16_type; break;
        case onnx::TensorProto::INT16: shape_type = shape::int16_type; break;
        case onnx::TensorProto::INT32: shape_type = shape::int32_type; break;
        case onnx::TensorProto::INT64: shape_type = shape::int64_type; break;
        case onnx::TensorProto::STRING:
            break; // throw std::runtime_error("Unsupported type STRING");
        case onnx::TensorProto::BOOL:
            break; // throw std::runtime_error("Unsupported type BOOL");
Paul's avatar
Paul committed
1358
        case onnx::TensorProto::FLOAT16: shape_type = shape::half_type; break;
Paul's avatar
Paul committed
1359
1360
1361
1362
1363
1364
1365
1366
1367
        case onnx::TensorProto::DOUBLE: shape_type = shape::double_type; break;
        case onnx::TensorProto::UINT32: shape_type = shape::uint32_type; break;
        case onnx::TensorProto::UINT64: shape_type = shape::uint64_type; break;
        case onnx::TensorProto::COMPLEX64:
            break; // throw std::runtime_error("Unsupported type COMPLEX64");
        case onnx::TensorProto::COMPLEX128:
            break; // throw std::runtime_error("Unsupported type COMPLEX128");
        }
        std::vector<std::size_t> dims;
Paul's avatar
Paul committed
1368
        auto&& tensor_dims = t.tensor_type().shape().dim();
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
        std::transform(tensor_dims.begin(),
                       tensor_dims.end(),
                       std::back_inserter(dims),
                       [](auto&& d) -> std::size_t {
                           if(not d.has_dim_value())
                           {
                               long default_batch_size = 1; // FIXME
                               return default_batch_size;
                           }
                           return d.dim_value();
                       });
Paul's avatar
Paul committed
1380
1381
        return {shape_type, dims};
    }
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403

    shape::type_t get_type(int dtype)
    {
        switch(dtype)
        {
        case 1: return shape::float_type;
        case 2: return shape::uint8_type;
        case 3: return shape::int8_type;
        case 4: return shape::uint16_type;
        case 5: return shape::int16_type;
        case 6: return shape::int32_type;
        case 7: return shape::int64_type;
        case 10: return shape::half_type;
        case 11: return shape::double_type;
        case 12: return shape::uint32_type;
        case 13: return shape::uint64_type;
        default:
        {
            MIGRAPHX_THROW("Prototensor data type " + std::to_string(dtype) + " not supported");
        }
        }
    }
Paul's avatar
Paul committed
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
};

program parse_onnx(const std::string& name)
{
    std::fstream input(name.c_str(), std::ios::in | std::ios::binary);
    onnx_parser parser;
#ifndef NDEBUG
    // Log the program when it can't be parsed
    try
    {
        parser.parse_from(input);
    }
    catch(...)
    {
        std::cerr << parser.prog << std::endl;
        throw;
    }
#else
    parser.parse_from(input);
#endif
    return std::move(parser.prog);
}

Paul's avatar
Paul committed
1427
} // namespace MIGRAPHX_INLINE_NS
Paul's avatar
Paul committed
1428
} // namespace migraphx