Tags: openvinotoolkit/openvino_tokenizers
Tags
Port Fixes from Master to Release (#315) * Fix Split Operation For Merging Modes (#311) * Fix Split Operation For Merging Modes * Format Tests * Update Stats * Update Stats (cherry picked from commit 45f441a) * Add max_length Option to CLI Convert Tool (#309) * Add max_length Option to CLI Convert Tool * Add max_length Option to CLI Convert Tool * Add max_length Option to CLI Convert Tool * Add max_length Option to CLI Convert Tool (cherry picked from commit 5c61c1a) * Update Regex For Clean Tokenization Spaces * Fix Replace Pattern Rewrite * Update Pass Rate * fixed sdl issue (#313) (cherry picked from commit c30086c) * Del Unused Regex --------- Co-authored-by: Mikhail Ryzhov <[email protected]>
Гse PCRE2 fallback for RegexNormalization' * add pcre2 * use string with length instead of ZERO_TERMINATION * add Licence to thirt party * set m_search_pattern_re to nullptr if PCRE2 is used * remove unnecessary const char* * fix typo in m_replace_pattern * Update Del Control Chars Regex * Update Pass Rate * fix win build * revert added_tests in conftest.py * revert Readme & pass rates * update README & pass_rates * do not log errors * add comment on malloc --------- Co-authored-by: Artur Paniukov <[email protected]>
Cherry Pick For 24.1 Release (#118) * Add Example RWKV to Readme (cherry picked from commit fb4142d) * Update Example (cherry picked from commit 6441bef) * Add Cache to VocabEncoder (cherry picked from commit fd32aa5) * Install torch CPU (cherry picked from commit dcc5fcd) * Parse QWEN EOD_TOKEN_ID (cherry picked from commit 029a341) * Fix Parse (cherry picked from commit 0298fea) * Fix Pass-Rate --------- Co-authored-by: Ilya Lavrenov <[email protected]>