Dev no emo (#123)
* Create emo_gen.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * update server.py, fix bugs in func get_text() and infer(). (#52) * Extract get_text() and infer() from webui.py. (#53) * Extract get_text() and infer() from webui.py. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * add emo emb * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * init emo gen * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * init emo * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * init emo * Delete bert/bert-base-japanese-v3 directory * Create .gitkeep * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Create add_punc.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix bug in bert_gen.py (#54) * Update README.md * fix bug in models.py (#56) * 更新 models.py * Fix japanese cleaner (#61) * 初步，睡觉明天继续写（ * 好好好放错分支了，熬夜是大忌 * [pre-commit.ci] pre-commit autoupdate (#55) * [pre-commit.ci] pre-commit autoupdate updates: - [github.com/pre-commit/pre-commit-hooks: v4.4.0 → v4.5.0](https://github.com/pre-commit/pre-commit-hooks/compare/v4.4.0...v4.5.0) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Create tokenizer_config.json * update preprocess_text.py：过滤一个音频匹配多个文本的情况 (#57) * update preprocess_text.py：过滤音频不存在的情况 (#58) * 修复日语cleaner和bert * better * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Sora <atri@suzakuintsubaki.com> * Apply Code Formatter Change * Add config.yml for global configuration. (#62) * Add config.yml for global configuration. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix bug in webui.py. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Rename config.yml to default_config.yml. Add ./config.yml to gitignore. * Add config.py to parse config.yml * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update webui.py (#65) * Update webui.py: 1. Add auto translation from Chinese to Japanese. 2. Start to use config.py in webui.py to set config instead of using the command line. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix (#68) * 加上ー * fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update infer.py and webui.py. Supports loading and inference models of 1.1.1 version. (#66) * Update infer.py and webui.py. Supports loading and inference models of 1.1.1 version. * Update config.json * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix bug in translate.py (#69) * Supports loading and inference models of 1.1、1.0.1、1.0 version. (#70) * Supports loading and inference models of 1.1、1.0.1、1.0 version. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Delete useless file in OldVersion --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update japanese.py (#71) Handling JA long pronunciations * 使用配置文件配置bert_gen.py, preprocess_text.py, resample.py (#72) * Update bert_gen.py, preprocess_text.py, resample.py. Support using config.yml in these files. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update bert_gen.py * Update bert_gen.py, fix bug. --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Delete bert/bert-base-japanese-v3 directory * Create config.json * Create tokenizer_config.json * Create vocab.txt * Update server.py. 支持多版本多模型 (#76) * Update server.py. 支持多版本多模型 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Dev webui (#77) * 申请pr (#75) * 2023/10/11 update 界面优化 * Update webui.py 翻译英文页面为中文 * Update train_ms.py 单卡训练 * 加入图片 * Update extern_subprocess.py * Update asr_transcript.py * Update asr_transcript.py * Update asr_transcript.py * Update extern_subprocess.py * Update asr_transcript.py * Update asr_transcript.py * Update asr_transcript.py * Update all_process.py * Update extern_subprocess.py * Update all_process.py * Update all_process.py * Update asr_transcript.py * Update extern_subprocess.py * Update webui.py * Create re_matching.py * Update webui.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update all_process.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update all_process.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update all_process.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update asr_transcript.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Pack 'update' functions into a module * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update all_process.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update asr_transcript.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update extern_subprocess.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update all_process.py * Update asr_transcript.py * Update webui.py * Add files via upload * Update extern_subprocess.py * Update all_process.py * Update asr_transcript.py * Update bert_gen.py * Update extern_subprocess.py * Update preprocess_text.py * Update re_matching.py * Update resample.py * Update update_status.py * Update update_status.py * Update webui.py * Update all_process.py * Update preprocess_text.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update train_ms.py --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> Co-authored-by: innnky <67028263+innnky@users.noreply.github.com> * Delete all_process.py * Delete asr_transcript.py * Delete extern_subprocess.py --------- Co-authored-by: spicysama <122108331+AnyaCoder@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: innnky <67028263+innnky@users.noreply.github.com> * Create config.json * Create preprocessor_config.json * Create vocab.json * Delete emotional/wav2vec2-large-robust-12-ft-emotion-msp-dim/.gitkeep * Update emo_gen.py * Delete add_punc.py * add emotion_clustering.i * Apply Code Formatter Change * Update models.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update preprocess_text.py (#78) * Update preprocess_text.py. 检测重复以及不存在的音频 (#79) * Handle Janpanese long pronunciations (#80) * Handle Janpanese long pronunciations * Update japanese.py * Update japanese.py * Use unified phonemes for Japanese long vowel (#82) * Use an unified phoneme for Japanese long vowel `symbol.py` has not been updated to ensure compatibility with older version models. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * 增加一个按钮，点击后可以按句子切分，添加“|” (#81) * Update re_matching.py * Update webui.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix phonemer bug (#83) * Fix phonemer bug * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix long vowel handler bug (#84) * Fix long vowel handler bug * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * 加入整合包管理器的特性：长文本合成可以自定义句间段间停顿 (#85) * Update webui.py * Update re_matching.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update train_ms.py * fix' * Update cleaner.py * add en * add en * Update english.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add en * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add en * add en * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add en * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * 更新 README.md * 更新 README.md * 更新 README.md * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Change phonemer to pyopenjtalk (#86) * Change phonemer to pyopenjtalk * 修改为openjtalk便于安装 --------- Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> * 更新 english.py * Fix english_bert_mock.py. (#87) * Add punctuation execptions (#88) * Add punctuation execptions * Ellipses exceptions * remove get bert * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix bug in oldVersion. (#89) * Update requirements.txt * change to large * rollback requirements.txt * Feat: Enable 1.1.1 models using fix-ver infer. (#91) * Feat: Enable 1.1.1 models using fix-ver infer. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Add Japanese accent (high-low) (#90) * Add punctuation execptions * Ellipses exceptions * Add Japanese accent * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Do not replace iteration mark (#92) * Add punctuation execptions * Ellipses exceptions * Add Japanese accent * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Do not replace iteration mark --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix: fix import error in oldVersion (#93) * Refactor: reusing model loading in webui.py and server.py. (#94) * Feat: Enable using config.yml in train_ms.py (#96) * 更新 emo_gen.py * Change emo_gen.py (#97) * Fix emo_gen bugs * Add multiprocess * Fix queue (#98) * Fix emo_gen bugs * Add multiprocess * Del var * Fix queue * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix training bugs (#99) * Updatge cluster notebook * Fix train * Fix filename * Update infer.py (#100) * Update infer.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Add reference audio (#101) * Add reference audio * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> * Fix: fix 1.1.1-fix (#102) * Fix infer bug (#103) * Feat: Add server_fastapi.py. (#104) * Feat: Add server_fastapi.py. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix: Update requirements.txt. --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix: requirements.txt. (#105) * Swith to deberta-v3-large (#106) * Swith to deberta-v3-large * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Feat: Update config.py. (#107) * Feat: Update config.py. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Dev fix (#108) * fix bugs when deploying * fix bugs when deploying * fix bugs when deploying * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Revert "Dev fix (#108)" (#109) This reverts commit 685e18a10498d602b1a9a26079340d11925646f0. * Dev fix (#110) * fix bugs when deploying * fix bugs when deploying * fix bugs when deploying * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix fixed bugs * fix fixed bugs * fix fixed bug 3 * fix fixed bug 4 * fix fixed bug 5 * fix * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Add emo vec quantizer (#111) Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> * Clean req and gitignore (#112) * Clean req and gitignore * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Switch to deberta-v2-large-japanese (#113) * Switch to deberta-v2-large-japanese * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix emo bugs (#114) * Fix english (#115) * Remove emo (#117) * Don't train codebook * Remove emo * Update * Update * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Merge dev into no-emo (#122) * [pre-commit.ci] pre-commit autoupdate (#95) * [pre-commit.ci] pre-commit autoupdate updates: - [github.com/astral-sh/ruff-pre-commit: v0.0.292 → v0.1.1](https://github.com/astral-sh/ruff-pre-commit/compare/v0.0.292...v0.1.1) - [github.com/psf/black: 23.9.1 → 23.10.0](https://github.com/psf/black/compare/23.9.1...23.10.0) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Don't train codebook (#116) * Update requirements.txt * Update english_bert_mock.py * Fix: server_fastapi.py (#118) * Fix: server_fastapi.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix: don't print debug logging. (#119) * Fix: don't print debug logging. * Feat: support emo_gen config * Fix config * Apply Code Formatter Change * 更新，修正bug (#121) * Feat: Update infer.py preprocess_text.py server_fastapi.py. * Fix resample.py. Maintain same directory structure in out_dir as in_dir. * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update resample.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update server_fastapi.py to no-emo ver * Update config.py, no emo config --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: OedoSoldier <31711261+OedoSoldier@users.noreply.github.com> Co-authored-by: Stardust·减 <star_dust_chen@foxmail.com> Co-authored-by: Stardust-minus <Stardust-minus@users.noreply.github.com> * Update train_ms.py * Update latest version info (#124) --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: jiangyuxiaoxiao <atri@suzakuintsubaki.com> Co-authored-by: AkitoLiu <39857739+Akito-UzukiP@users.noreply.github.com> Co-authored-by: Stardust-minus <Stardust-minus@users.noreply.github.com> Co-authored-by: OedoSoldier <31711261+OedoSoldier@users.noreply.github.com> Co-authored-by: spicysama <122108331+AnyaCoder@users.noreply.github.com> Co-authored-by: innnky <67028263+innnky@users.noreply.github.com> Co-authored-by: YYuX-1145 <138500330+YYuX-1145@users.noreply.github.com>
This commit is contained in:
Stardust·减
34
bert/bert-base-japanese-v3/.gitattributes
vendored
Normal file
34
bert/bert-base-japanese-v3/.gitattributes
vendored
Normal file
@@ -0,0 +1,34 @@
|
||||
*.7z filter=lfs diff=lfs merge=lfs -text
|
||||
*.arrow filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin filter=lfs diff=lfs merge=lfs -text
|
||||
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
||||
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
||||
*.ftz filter=lfs diff=lfs merge=lfs -text
|
||||
*.gz filter=lfs diff=lfs merge=lfs -text
|
||||
*.h5 filter=lfs diff=lfs merge=lfs -text
|
||||
*.joblib filter=lfs diff=lfs merge=lfs -text
|
||||
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
||||
*.model filter=lfs diff=lfs merge=lfs -text
|
||||
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
||||
*.npy filter=lfs diff=lfs merge=lfs -text
|
||||
*.npz filter=lfs diff=lfs merge=lfs -text
|
||||
*.onnx filter=lfs diff=lfs merge=lfs -text
|
||||
*.ot filter=lfs diff=lfs merge=lfs -text
|
||||
*.parquet filter=lfs diff=lfs merge=lfs -text
|
||||
*.pb filter=lfs diff=lfs merge=lfs -text
|
||||
*.pickle filter=lfs diff=lfs merge=lfs -text
|
||||
*.pkl filter=lfs diff=lfs merge=lfs -text
|
||||
*.pt filter=lfs diff=lfs merge=lfs -text
|
||||
*.pth filter=lfs diff=lfs merge=lfs -text
|
||||
*.rar filter=lfs diff=lfs merge=lfs -text
|
||||
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
||||
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tflite filter=lfs diff=lfs merge=lfs -text
|
||||
*.tgz filter=lfs diff=lfs merge=lfs -text
|
||||
*.wasm filter=lfs diff=lfs merge=lfs -text
|
||||
*.xz filter=lfs diff=lfs merge=lfs -text
|
||||
*.zip filter=lfs diff=lfs merge=lfs -text
|
||||
*.zst filter=lfs diff=lfs merge=lfs -text
|
||||
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
||||
34
bert/bert-large-japanese-v2/.gitattributes
vendored
Normal file
34
bert/bert-large-japanese-v2/.gitattributes
vendored
Normal file
@@ -0,0 +1,34 @@
|
||||
*.7z filter=lfs diff=lfs merge=lfs -text
|
||||
*.arrow filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin filter=lfs diff=lfs merge=lfs -text
|
||||
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
||||
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
||||
*.ftz filter=lfs diff=lfs merge=lfs -text
|
||||
*.gz filter=lfs diff=lfs merge=lfs -text
|
||||
*.h5 filter=lfs diff=lfs merge=lfs -text
|
||||
*.joblib filter=lfs diff=lfs merge=lfs -text
|
||||
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
||||
*.model filter=lfs diff=lfs merge=lfs -text
|
||||
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
||||
*.npy filter=lfs diff=lfs merge=lfs -text
|
||||
*.npz filter=lfs diff=lfs merge=lfs -text
|
||||
*.onnx filter=lfs diff=lfs merge=lfs -text
|
||||
*.ot filter=lfs diff=lfs merge=lfs -text
|
||||
*.parquet filter=lfs diff=lfs merge=lfs -text
|
||||
*.pb filter=lfs diff=lfs merge=lfs -text
|
||||
*.pickle filter=lfs diff=lfs merge=lfs -text
|
||||
*.pkl filter=lfs diff=lfs merge=lfs -text
|
||||
*.pt filter=lfs diff=lfs merge=lfs -text
|
||||
*.pth filter=lfs diff=lfs merge=lfs -text
|
||||
*.rar filter=lfs diff=lfs merge=lfs -text
|
||||
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
||||
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tflite filter=lfs diff=lfs merge=lfs -text
|
||||
*.tgz filter=lfs diff=lfs merge=lfs -text
|
||||
*.wasm filter=lfs diff=lfs merge=lfs -text
|
||||
*.xz filter=lfs diff=lfs merge=lfs -text
|
||||
*.zip filter=lfs diff=lfs merge=lfs -text
|
||||
*.zst filter=lfs diff=lfs merge=lfs -text
|
||||
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
||||
53
bert/bert-large-japanese-v2/README.md
Normal file
53
bert/bert-large-japanese-v2/README.md
Normal file
@@ -0,0 +1,53 @@
|
||||
---
|
||||
license: apache-2.0
|
||||
datasets:
|
||||
- cc100
|
||||
- wikipedia
|
||||
language:
|
||||
- ja
|
||||
widget:
|
||||
- text: 東北大学で[MASK]の研究をしています。
|
||||
---
|
||||
|
||||
# BERT large Japanese (unidic-lite with whole word masking, CC-100 and jawiki-20230102)
|
||||
|
||||
This is a [BERT](https://github.com/google-research/bert) model pretrained on texts in the Japanese language.
|
||||
|
||||
This version of the model processes input texts with word-level tokenization based on the Unidic 2.1.2 dictionary (available in [unidic-lite](https://pypi.org/project/unidic-lite/) package), followed by the WordPiece subword tokenization.
|
||||
Additionally, the model is trained with the whole word masking enabled for the masked language modeling (MLM) objective.
|
||||
|
||||
The codes for the pretraining are available at [cl-tohoku/bert-japanese](https://github.com/cl-tohoku/bert-japanese/).
|
||||
|
||||
## Model architecture
|
||||
|
||||
The model architecture is the same as the original BERT large model; 24 layers, 1024 dimensions of hidden states, and 16 attention heads.
|
||||
|
||||
## Training Data
|
||||
|
||||
The model is trained on the Japanese portion of [CC-100 dataset](https://data.statmt.org/cc-100/) and the Japanese version of Wikipedia.
|
||||
For Wikipedia, we generated a text corpus from the [Wikipedia Cirrussearch dump file](https://dumps.wikimedia.org/other/cirrussearch/) as of January 2, 2023.
|
||||
The corpus files generated from CC-100 and Wikipedia are 74.3GB and 4.9GB in size and consist of approximately 392M and 34M sentences, respectively.
|
||||
|
||||
For the purpose of splitting texts into sentences, we used [fugashi](https://github.com/polm/fugashi) with [mecab-ipadic-NEologd](https://github.com/neologd/mecab-ipadic-neologd) dictionary (v0.0.7).
|
||||
|
||||
## Tokenization
|
||||
|
||||
The texts are first tokenized by MeCab with the Unidic 2.1.2 dictionary and then split into subwords by the WordPiece algorithm.
|
||||
The vocabulary size is 32768.
|
||||
|
||||
We used [fugashi](https://github.com/polm/fugashi) and [unidic-lite](https://github.com/polm/unidic-lite) packages for the tokenization.
|
||||
|
||||
## Training
|
||||
|
||||
We trained the model first on the CC-100 corpus for 1M steps and then on the Wikipedia corpus for another 1M steps.
|
||||
For training of the MLM (masked language modeling) objective, we introduced whole word masking in which all of the subword tokens corresponding to a single word (tokenized by MeCab) are masked at once.
|
||||
|
||||
For training of each model, we used a v3-8 instance of Cloud TPUs provided by [TPU Research Cloud](https://sites.research.google/trc/about/).
|
||||
|
||||
## Licenses
|
||||
|
||||
The pretrained models are distributed under the Apache License 2.0.
|
||||
|
||||
## Acknowledgments
|
||||
|
||||
This model is trained with Cloud TPUs provided by [TPU Research Cloud](https://sites.research.google/trc/about/) program.
|
||||
19
bert/bert-large-japanese-v2/config.json
Normal file
19
bert/bert-large-japanese-v2/config.json
Normal file
@@ -0,0 +1,19 @@
|
||||
{
|
||||
"architectures": [
|
||||
"BertForPreTraining"
|
||||
],
|
||||
"attention_probs_dropout_prob": 0.1,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_dropout_prob": 0.1,
|
||||
"hidden_size": 1024,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 4096,
|
||||
"layer_norm_eps": 1e-12,
|
||||
"max_position_embeddings": 512,
|
||||
"model_type": "bert",
|
||||
"num_attention_heads": 16,
|
||||
"num_hidden_layers": 24,
|
||||
"pad_token_id": 0,
|
||||
"type_vocab_size": 2,
|
||||
"vocab_size": 32768
|
||||
}
|
||||
10
bert/bert-large-japanese-v2/tokenizer_config.json
Normal file
10
bert/bert-large-japanese-v2/tokenizer_config.json
Normal file
@@ -0,0 +1,10 @@
|
||||
{
|
||||
"tokenizer_class": "BertJapaneseTokenizer",
|
||||
"model_max_length": 512,
|
||||
"do_lower_case": false,
|
||||
"word_tokenizer_type": "mecab",
|
||||
"subword_tokenizer_type": "wordpiece",
|
||||
"mecab_kwargs": {
|
||||
"mecab_dic": "unidic_lite"
|
||||
}
|
||||
}
|
||||
32768
bert/bert-large-japanese-v2/vocab.txt
Normal file
32768
bert/bert-large-japanese-v2/vocab.txt
Normal file
File diff suppressed because it is too large
Load Diff
@@ -1 +0,0 @@
|
||||
*.bin
|
||||
34
bert/deberta-v2-large-japanese/.gitattributes
vendored
Normal file
34
bert/deberta-v2-large-japanese/.gitattributes
vendored
Normal file
@@ -0,0 +1,34 @@
|
||||
*.7z filter=lfs diff=lfs merge=lfs -text
|
||||
*.arrow filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin filter=lfs diff=lfs merge=lfs -text
|
||||
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
||||
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
||||
*.ftz filter=lfs diff=lfs merge=lfs -text
|
||||
*.gz filter=lfs diff=lfs merge=lfs -text
|
||||
*.h5 filter=lfs diff=lfs merge=lfs -text
|
||||
*.joblib filter=lfs diff=lfs merge=lfs -text
|
||||
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
||||
*.model filter=lfs diff=lfs merge=lfs -text
|
||||
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
||||
*.npy filter=lfs diff=lfs merge=lfs -text
|
||||
*.npz filter=lfs diff=lfs merge=lfs -text
|
||||
*.onnx filter=lfs diff=lfs merge=lfs -text
|
||||
*.ot filter=lfs diff=lfs merge=lfs -text
|
||||
*.parquet filter=lfs diff=lfs merge=lfs -text
|
||||
*.pb filter=lfs diff=lfs merge=lfs -text
|
||||
*.pickle filter=lfs diff=lfs merge=lfs -text
|
||||
*.pkl filter=lfs diff=lfs merge=lfs -text
|
||||
*.pt filter=lfs diff=lfs merge=lfs -text
|
||||
*.pth filter=lfs diff=lfs merge=lfs -text
|
||||
*.rar filter=lfs diff=lfs merge=lfs -text
|
||||
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
||||
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tflite filter=lfs diff=lfs merge=lfs -text
|
||||
*.tgz filter=lfs diff=lfs merge=lfs -text
|
||||
*.wasm filter=lfs diff=lfs merge=lfs -text
|
||||
*.xz filter=lfs diff=lfs merge=lfs -text
|
||||
*.zip filter=lfs diff=lfs merge=lfs -text
|
||||
*.zst filter=lfs diff=lfs merge=lfs -text
|
||||
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
||||
111
bert/deberta-v2-large-japanese/README.md
Normal file
111
bert/deberta-v2-large-japanese/README.md
Normal file
@@ -0,0 +1,111 @@
|
||||
---
|
||||
language: ja
|
||||
license: cc-by-sa-4.0
|
||||
library_name: transformers
|
||||
tags:
|
||||
- deberta
|
||||
- deberta-v2
|
||||
- fill-mask
|
||||
datasets:
|
||||
- wikipedia
|
||||
- cc100
|
||||
- oscar
|
||||
metrics:
|
||||
- accuracy
|
||||
mask_token: "[MASK]"
|
||||
widget:
|
||||
- text: "京都 大学 で 自然 言語 処理 を [MASK] する 。"
|
||||
---
|
||||
|
||||
# Model Card for Japanese DeBERTa V2 large
|
||||
|
||||
## Model description
|
||||
|
||||
This is a Japanese DeBERTa V2 large model pre-trained on Japanese Wikipedia, the Japanese portion of CC-100, and the
|
||||
Japanese portion of OSCAR.
|
||||
|
||||
## How to use
|
||||
|
||||
You can use this model for masked language modeling as follows:
|
||||
|
||||
```python
|
||||
from transformers import AutoTokenizer, AutoModelForMaskedLM
|
||||
|
||||
tokenizer = AutoTokenizer.from_pretrained('ku-nlp/deberta-v2-large-japanese')
|
||||
model = AutoModelForMaskedLM.from_pretrained('ku-nlp/deberta-v2-large-japanese')
|
||||
|
||||
sentence = '京都 大学 で 自然 言語 処理 を [MASK] する 。' # input should be segmented into words by Juman++ in advance
|
||||
encoding = tokenizer(sentence, return_tensors='pt')
|
||||
...
|
||||
```
|
||||
|
||||
You can also fine-tune this model on downstream tasks.
|
||||
|
||||
## Tokenization
|
||||
|
||||
The input text should be segmented into words by [Juman++](https://github.com/ku-nlp/jumanpp) in
|
||||
advance. [Juman++ 2.0.0-rc3](https://github.com/ku-nlp/jumanpp/releases/tag/v2.0.0-rc3) was used for pre-training. Each
|
||||
word is tokenized into subwords by [sentencepiece](https://github.com/google/sentencepiece).
|
||||
|
||||
## Training data
|
||||
|
||||
We used the following corpora for pre-training:
|
||||
|
||||
- Japanese Wikipedia (as of 20221020, 3.2GB, 27M sentences, 1.3M documents)
|
||||
- Japanese portion of CC-100 (85GB, 619M sentences, 66M documents)
|
||||
- Japanese portion of OSCAR (54GB, 326M sentences, 25M documents)
|
||||
|
||||
Note that we filtered out documents annotated with "header", "footer", or "noisy" tags in OSCAR.
|
||||
Also note that Japanese Wikipedia was duplicated 10 times to make the total size of the corpus comparable to that of
|
||||
CC-100 and OSCAR. As a result, the total size of the training data is 171GB.
|
||||
|
||||
## Training procedure
|
||||
|
||||
We first segmented texts in the corpora into words using [Juman++](https://github.com/ku-nlp/jumanpp).
|
||||
Then, we built a sentencepiece model with 32000 tokens including words ([JumanDIC](https://github.com/ku-nlp/JumanDIC))
|
||||
and subwords induced by the unigram language model of [sentencepiece](https://github.com/google/sentencepiece).
|
||||
|
||||
We tokenized the segmented corpora into subwords using the sentencepiece model and trained the Japanese DeBERTa model
|
||||
using [transformers](https://github.com/huggingface/transformers) library.
|
||||
The training took 36 days using 8 NVIDIA A100-SXM4-40GB GPUs.
|
||||
|
||||
The following hyperparameters were used during pre-training:
|
||||
|
||||
- learning_rate: 1e-4
|
||||
- per_device_train_batch_size: 18
|
||||
- distributed_type: multi-GPU
|
||||
- num_devices: 8
|
||||
- gradient_accumulation_steps: 16
|
||||
- total_train_batch_size: 2,304
|
||||
- max_seq_length: 512
|
||||
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-06
|
||||
- lr_scheduler_type: linear schedule with warmup
|
||||
- training_steps: 300,000
|
||||
- warmup_steps: 10,000
|
||||
|
||||
The accuracy of the trained model on the masked language modeling task was 0.799.
|
||||
The evaluation set consists of 5,000 randomly sampled documents from each of the training corpora.
|
||||
|
||||
## Fine-tuning on NLU tasks
|
||||
|
||||
We fine-tuned the following models and evaluated them on the dev set of JGLUE.
|
||||
We tuned learning rate and training epochs for each model and task
|
||||
following [the JGLUE paper](https://www.jstage.jst.go.jp/article/jnlp/30/1/30_63/_pdf/-char/ja).
|
||||
|
||||
| Model | MARC-ja/acc | JSTS/pearson | JSTS/spearman | JNLI/acc | JSQuAD/EM | JSQuAD/F1 | JComQA/acc |
|
||||
|-------------------------------|-------------|--------------|---------------|----------|-----------|-----------|------------|
|
||||
| Waseda RoBERTa base | 0.965 | 0.913 | 0.876 | 0.905 | 0.853 | 0.916 | 0.853 |
|
||||
| Waseda RoBERTa large (seq512) | 0.969 | 0.925 | 0.890 | 0.928 | 0.910 | 0.955 | 0.900 |
|
||||
| LUKE Japanese base* | 0.965 | 0.916 | 0.877 | 0.912 | - | - | 0.842 |
|
||||
| LUKE Japanese large* | 0.965 | 0.932 | 0.902 | 0.927 | - | - | 0.893 |
|
||||
| DeBERTaV2 base | 0.970 | 0.922 | 0.886 | 0.922 | 0.899 | 0.951 | 0.873 |
|
||||
| DeBERTaV2 large | 0.968 | 0.925 | 0.892 | 0.924 | 0.912 | 0.959 | 0.890 |
|
||||
|
||||
*The scores of LUKE are from [the official repository](https://github.com/studio-ousia/luke).
|
||||
|
||||
## Acknowledgments
|
||||
|
||||
This work was supported by Joint Usage/Research Center for Interdisciplinary Large-scale Information Infrastructures (
|
||||
JHPCN) through General Collaboration Project no. jh221004, "Developing a Platform for Constructing and Sharing of
|
||||
Large-Scale Japanese Language Models".
|
||||
For training models, we used the mdx: a platform for the data-driven future.
|
||||
38
bert/deberta-v2-large-japanese/config.json
Normal file
38
bert/deberta-v2-large-japanese/config.json
Normal file
@@ -0,0 +1,38 @@
|
||||
{
|
||||
"_name_or_path": "configs/deberta_v2_large.json",
|
||||
"architectures": [
|
||||
"DebertaV2ForMaskedLM"
|
||||
],
|
||||
"attention_head_size": 64,
|
||||
"attention_probs_dropout_prob": 0.1,
|
||||
"conv_act": "gelu",
|
||||
"conv_kernel_size": 3,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_dropout_prob": 0.1,
|
||||
"hidden_size": 1024,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 4096,
|
||||
"layer_norm_eps": 1e-07,
|
||||
"max_position_embeddings": 512,
|
||||
"max_relative_positions": -1,
|
||||
"model_type": "deberta-v2",
|
||||
"norm_rel_ebd": "layer_norm",
|
||||
"num_attention_heads": 16,
|
||||
"num_hidden_layers": 24,
|
||||
"pad_token_id": 0,
|
||||
"pooler_dropout": 0,
|
||||
"pooler_hidden_act": "gelu",
|
||||
"pooler_hidden_size": 1024,
|
||||
"pos_att_type": [
|
||||
"p2c",
|
||||
"c2p"
|
||||
],
|
||||
"position_biased_input": false,
|
||||
"position_buckets": 256,
|
||||
"relative_attention": true,
|
||||
"share_att_key": true,
|
||||
"torch_dtype": "float32",
|
||||
"transformers_version": "4.23.1",
|
||||
"type_vocab_size": 0,
|
||||
"vocab_size": 32000
|
||||
}
|
||||
9
bert/deberta-v2-large-japanese/special_tokens_map.json
Normal file
9
bert/deberta-v2-large-japanese/special_tokens_map.json
Normal file
@@ -0,0 +1,9 @@
|
||||
{
|
||||
"bos_token": "[CLS]",
|
||||
"cls_token": "[CLS]",
|
||||
"eos_token": "[SEP]",
|
||||
"mask_token": "[MASK]",
|
||||
"pad_token": "[PAD]",
|
||||
"sep_token": "[SEP]",
|
||||
"unk_token": "[UNK]"
|
||||
}
|
||||
128169
bert/deberta-v2-large-japanese/tokenizer.json
Normal file
128169
bert/deberta-v2-large-japanese/tokenizer.json
Normal file
File diff suppressed because one or more lines are too long
15
bert/deberta-v2-large-japanese/tokenizer_config.json
Normal file
15
bert/deberta-v2-large-japanese/tokenizer_config.json
Normal file
@@ -0,0 +1,15 @@
|
||||
{
|
||||
"bos_token": "[CLS]",
|
||||
"cls_token": "[CLS]",
|
||||
"do_lower_case": false,
|
||||
"eos_token": "[SEP]",
|
||||
"keep_accents": true,
|
||||
"mask_token": "[MASK]",
|
||||
"pad_token": "[PAD]",
|
||||
"sep_token": "[SEP]",
|
||||
"sp_model_kwargs": {},
|
||||
"special_tokens_map_file": null,
|
||||
"split_by_punct": false,
|
||||
"tokenizer_class": "DebertaV2Tokenizer",
|
||||
"unk_token": "[UNK]"
|
||||
}
|
||||
27
bert/deberta-v3-large/.gitattributes
vendored
Normal file
27
bert/deberta-v3-large/.gitattributes
vendored
Normal file
@@ -0,0 +1,27 @@
|
||||
*.7z filter=lfs diff=lfs merge=lfs -text
|
||||
*.arrow filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin filter=lfs diff=lfs merge=lfs -text
|
||||
*.bin.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
||||
*.ftz filter=lfs diff=lfs merge=lfs -text
|
||||
*.gz filter=lfs diff=lfs merge=lfs -text
|
||||
*.h5 filter=lfs diff=lfs merge=lfs -text
|
||||
*.joblib filter=lfs diff=lfs merge=lfs -text
|
||||
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.model filter=lfs diff=lfs merge=lfs -text
|
||||
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
||||
*.onnx filter=lfs diff=lfs merge=lfs -text
|
||||
*.ot filter=lfs diff=lfs merge=lfs -text
|
||||
*.parquet filter=lfs diff=lfs merge=lfs -text
|
||||
*.pb filter=lfs diff=lfs merge=lfs -text
|
||||
*.pt filter=lfs diff=lfs merge=lfs -text
|
||||
*.pth filter=lfs diff=lfs merge=lfs -text
|
||||
*.rar filter=lfs diff=lfs merge=lfs -text
|
||||
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
||||
*.tflite filter=lfs diff=lfs merge=lfs -text
|
||||
*.tgz filter=lfs diff=lfs merge=lfs -text
|
||||
*.xz filter=lfs diff=lfs merge=lfs -text
|
||||
*.zip filter=lfs diff=lfs merge=lfs -text
|
||||
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
||||
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
||||
93
bert/deberta-v3-large/README.md
Normal file
93
bert/deberta-v3-large/README.md
Normal file
@@ -0,0 +1,93 @@
|
||||
---
|
||||
language: en
|
||||
tags:
|
||||
- deberta
|
||||
- deberta-v3
|
||||
- fill-mask
|
||||
thumbnail: https://huggingface.co/front/thumbnails/microsoft.png
|
||||
license: mit
|
||||
---
|
||||
|
||||
## DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing
|
||||
|
||||
[DeBERTa](https://arxiv.org/abs/2006.03654) improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data.
|
||||
|
||||
In [DeBERTa V3](https://arxiv.org/abs/2111.09543), we further improved the efficiency of DeBERTa using ELECTRA-Style pre-training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our [paper](https://arxiv.org/abs/2111.09543).
|
||||
|
||||
Please check the [official repository](https://github.com/microsoft/DeBERTa) for more implementation details and updates.
|
||||
|
||||
The DeBERTa V3 large model comes with 24 layers and a hidden size of 1024. It has 304M backbone parameters with a vocabulary containing 128K tokens which introduces 131M parameters in the Embedding layer. This model was trained using the 160GB data as DeBERTa V2.
|
||||
|
||||
|
||||
#### Fine-tuning on NLU tasks
|
||||
|
||||
We present the dev results on SQuAD 2.0 and MNLI tasks.
|
||||
|
||||
| Model |Vocabulary(K)|Backbone #Params(M)| SQuAD 2.0(F1/EM) | MNLI-m/mm(ACC)|
|
||||
|-------------------|----------|-------------------|-----------|----------|
|
||||
| RoBERTa-large |50 |304 | 89.4/86.5 | 90.2 |
|
||||
| XLNet-large |32 |- | 90.6/87.9 | 90.8 |
|
||||
| DeBERTa-large |50 |- | 90.7/88.0 | 91.3 |
|
||||
| **DeBERTa-v3-large**|128|304 | **91.5/89.0**| **91.8/91.9**|
|
||||
|
||||
|
||||
#### Fine-tuning with HF transformers
|
||||
|
||||
```bash
|
||||
#!/bin/bash
|
||||
|
||||
cd transformers/examples/pytorch/text-classification/
|
||||
|
||||
pip install datasets
|
||||
export TASK_NAME=mnli
|
||||
|
||||
output_dir="ds_results"
|
||||
|
||||
num_gpus=8
|
||||
|
||||
batch_size=8
|
||||
|
||||
python -m torch.distributed.launch --nproc_per_node=${num_gpus} \
|
||||
run_glue.py \
|
||||
--model_name_or_path microsoft/deberta-v3-large \
|
||||
--task_name $TASK_NAME \
|
||||
--do_train \
|
||||
--do_eval \
|
||||
--evaluation_strategy steps \
|
||||
--max_seq_length 256 \
|
||||
--warmup_steps 50 \
|
||||
--per_device_train_batch_size ${batch_size} \
|
||||
--learning_rate 6e-6 \
|
||||
--num_train_epochs 2 \
|
||||
--output_dir $output_dir \
|
||||
--overwrite_output_dir \
|
||||
--logging_steps 1000 \
|
||||
--logging_dir $output_dir
|
||||
|
||||
```
|
||||
|
||||
### Citation
|
||||
|
||||
If you find DeBERTa useful for your work, please cite the following papers:
|
||||
|
||||
``` latex
|
||||
@misc{he2021debertav3,
|
||||
title={DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing},
|
||||
author={Pengcheng He and Jianfeng Gao and Weizhu Chen},
|
||||
year={2021},
|
||||
eprint={2111.09543},
|
||||
archivePrefix={arXiv},
|
||||
primaryClass={cs.CL}
|
||||
}
|
||||
```
|
||||
|
||||
``` latex
|
||||
@inproceedings{
|
||||
he2021deberta,
|
||||
title={DEBERTA: DECODING-ENHANCED BERT WITH DISENTANGLED ATTENTION},
|
||||
author={Pengcheng He and Xiaodong Liu and Jianfeng Gao and Weizhu Chen},
|
||||
booktitle={International Conference on Learning Representations},
|
||||
year={2021},
|
||||
url={https://openreview.net/forum?id=XPZIaotutsD}
|
||||
}
|
||||
```
|
||||
22
bert/deberta-v3-large/config.json
Normal file
22
bert/deberta-v3-large/config.json
Normal file
@@ -0,0 +1,22 @@
|
||||
{
|
||||
"model_type": "deberta-v2",
|
||||
"attention_probs_dropout_prob": 0.1,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_dropout_prob": 0.1,
|
||||
"hidden_size": 1024,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 4096,
|
||||
"max_position_embeddings": 512,
|
||||
"relative_attention": true,
|
||||
"position_buckets": 256,
|
||||
"norm_rel_ebd": "layer_norm",
|
||||
"share_att_key": true,
|
||||
"pos_att_type": "p2c|c2p",
|
||||
"layer_norm_eps": 1e-7,
|
||||
"max_relative_positions": -1,
|
||||
"position_biased_input": false,
|
||||
"num_attention_heads": 16,
|
||||
"num_hidden_layers": 24,
|
||||
"type_vocab_size": 0,
|
||||
"vocab_size": 128100
|
||||
}
|
||||
22
bert/deberta-v3-large/generator_config.json
Normal file
22
bert/deberta-v3-large/generator_config.json
Normal file
@@ -0,0 +1,22 @@
|
||||
{
|
||||
"model_type": "deberta-v2",
|
||||
"attention_probs_dropout_prob": 0.1,
|
||||
"hidden_act": "gelu",
|
||||
"hidden_dropout_prob": 0.1,
|
||||
"hidden_size": 1024,
|
||||
"initializer_range": 0.02,
|
||||
"intermediate_size": 4096,
|
||||
"max_position_embeddings": 512,
|
||||
"relative_attention": true,
|
||||
"position_buckets": 256,
|
||||
"norm_rel_ebd": "layer_norm",
|
||||
"share_att_key": true,
|
||||
"pos_att_type": "p2c|c2p",
|
||||
"layer_norm_eps": 1e-7,
|
||||
"max_relative_positions": -1,
|
||||
"position_biased_input": false,
|
||||
"num_attention_heads": 16,
|
||||
"num_hidden_layers": 12,
|
||||
"type_vocab_size": 0,
|
||||
"vocab_size": 128100
|
||||
}
|
||||
4
bert/deberta-v3-large/tokenizer_config.json
Normal file
4
bert/deberta-v3-large/tokenizer_config.json
Normal file
@@ -0,0 +1,4 @@
|
||||
{
|
||||
"do_lower_case": false,
|
||||
"vocab_type": "spm"
|
||||
}
|
||||
Reference in New Issue
Block a user