microsoft/onnxruntime-extensions
Publicmirrored fromhttps://github.com/microsoft/onnxruntime-extensionsAvailable
operators/tokenizer/basic_tokenizer.hpp
38lines · modecode
| 1 | // Copyright (c) Microsoft Corporation. All rights reserved. |
| 2 | // Licensed under the MIT License. |
| 3 | |
| 4 | #pragma once |
| 5 | |
| 6 | #include "ocos.h" |
| 7 | #include "string_utils.h" |
| 8 | #include "ustring.h" |
| 9 | |
| 10 | class BasicTokenizer { |
| 11 | public: |
| 12 | BasicTokenizer(bool do_lower_case, bool tokenize_chinese_chars, bool strip_accents, bool tokenize_punctuation, |
| 13 | bool remove_control_chars); |
| 14 | std::vector<ustring> Tokenize(ustring text); |
| 15 | |
| 16 | private: |
| 17 | bool do_lower_case_; |
| 18 | bool strip_accents_; |
| 19 | bool tokenize_chinese_chars_; |
| 20 | bool tokenize_punctuation_; |
| 21 | bool remove_control_chars_; |
| 22 | }; |
| 23 | |
| 24 | struct KernelBasicTokenizer : BaseKernel { |
| 25 | KernelBasicTokenizer(const OrtApi& api, const OrtKernelInfo& info); |
| 26 | void Compute(OrtKernelContext* context); |
| 27 | |
| 28 | private: |
| 29 | std::shared_ptr<BasicTokenizer> tokenizer_; |
| 30 | }; |
| 31 | |
| 32 | struct CustomOpBasicTokenizer : OrtW::CustomOpBase<CustomOpBasicTokenizer, KernelBasicTokenizer> { |
| 33 | const char* GetName() const; |
| 34 | size_t GetInputTypeCount() const; |
| 35 | ONNXTensorElementDataType GetInputType(size_t index) const; |
| 36 | size_t GetOutputTypeCount() const; |
| 37 | ONNXTensorElementDataType GetOutputType(size_t index) const; |
| 38 | }; |
| 39 | |