/* * The MIT License (MIT) * * Copyright (c) 2015-2022 Advanced Micro Devices, Inc. All rights reserved. * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to deal * in the Software without restriction, including without limitation the rights * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell * copies of the Software, and to permit persons to whom the Software is * furnished to do so, subject to the following conditions: * * The above copyright notice and this permission notice shall be included in * all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN * THE SOFTWARE. */ #include #include #include #include #include #include namespace migraphx { inline namespace MIGRAPHX_INLINE_NS { namespace onnx { struct parse_splitToSequence : op_parser { std::vector operators() const { return {{"SplitToSequence"}}; } std::vector parse(const op_desc& opd, const onnx_parser& parser, onnx_parser::node_info info, std::vector args) const { int64_t axis = 0; int64_t keep_dims = 1; if(contains(info.attributes, "axis")) { axis = parser.parse_value(info.attributes.at("axis")).at(); } if(contains(info.attributes, "keepdims")) { keep_dims = parser.parse_value(info.attributes.at("keepdims")).at(); } auto lens = args[0]->get_shape().lens(); int64_t n_rank = lens.size(); int64_t tuned_axis = tune_axis(n_rank, axis, opd.op_name); std::vector vec_splits; if(args.size() == 2) { auto s = args[1]->eval(); check_arg_empty(s, "SplitToSequence: dynamic shape is not supported"); const auto split_shape = s.get_shape(); // check all split args > 1 for(const auto split_arg : split_shape.lens()) { assert(split_arg > 0); } if(split_shape.scalar()) { // Split equally along one axis based on desired split auto split_output = lens[tuned_axis] / split_shape.lens().at(0); auto dl = lens[tuned_axis] / info.num_outputs; vec_splits.resize(split_output, dl); } else { s.visit([&](auto v) { vec_splits.assign(v.begin(), v.end()); }); } if(std::accumulate(vec_splits.begin(), vec_splits.end(), int64_t(0)) != static_cast(lens[tuned_axis])) { MIGRAPHX_THROW("PARSE_SPLIT_TO_SEQ: sum of split attribute unequal to dim size of " "axis! Axis " + std::to_string(lens[tuned_axis]) + " Split " + to_string_range(vec_splits)); } } else { if(keep_dims == 0) { if((lens[tuned_axis] % info.num_outputs) != 0) { MIGRAPHX_THROW("PARSE_SPLIT_TO_SEQ: input cannot be equally divided into " + std::to_string(info.num_outputs) + " splits!"); } auto dl = lens[tuned_axis] / info.num_outputs; vec_splits.resize(info.num_outputs, dl); } else { vec_splits.resize(info.num_outputs, lens[tuned_axis]); } } std::vector ret_ins; int64_t start = 0; for(auto sl : vec_splits) { ret_ins.push_back(info.add_instruction( make_op("slice", {{"axes", {axis}}, {"starts", {start}}, {"ends", {start + sl}}}), args[0])); start += sl; } return ret_ins; } }; } // namespace onnx } // namespace MIGRAPHX_INLINE_NS } // namespace migraphx