Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[TTS] porting VITS implementation (NVIDIA#5600)
* Disable loss typecheck * Fix spectrogram lengths * Remove Precision 16 requirement * Address lgtm alerts * clean up unused code * Address lgtm alerts * Refactor audio_to_mel_torch method * Use NeMo FilterBank to get melspec Todo: set self.fb * Fix filterbank max frequency to match with original VITS * Fix filterbank features correct length * Address lgtm issues * Remove print statements * Remove stft_pad_amount * new structure for tts datasets in script folder Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * remove cmudict downloading Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * rename mixertts dataset, add vocoder dataset Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * add libritts processing Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * update tts dataset and libritts get data Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * fix bugs in vocoder ds Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * add ds * changed vits yaml * rm yaml * fix yaml and model * Added scaler * refactored yaml * managed to run in fp16 * refactoring Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * fix small bugs and add new todos Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * fix optimizers Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * Port Variational Inference with Adversarial Learning (VITS) to NeMo TTS (NVIDIA#6) * Add vits files Add vits_losses.py, vits_modules.py and vits.py. * Move non-vits models to modules * Add vits.yaml * Add _loader to vits.py * Add basic template for vits * Update vits.yaml with vits parameters * Remove extra space * Add top level training script * Add some variables to vits yaml * Add forward and training methods * Fix imports * Added validation step * Log training losses * Update loss calls to use class attributes * Add VITS to models list * Fix all imports * Remove old module calls * Fix typo in monotonic align import * Modified validation step 1. reverted to tensorboard 2. validation_step logs audio, mel-spec for batch 0 3. validation_step_alt logs audio, mel-spec for batch 0 and loss_mel * Fix imports for VITS * Remove old module calls * Fix typo in monotonic align import * Modified validation step 1. reverted to tensorboard 2. validation_step logs audio, mel-spec for batch 0 3. validation_step_alt logs audio, mel-spec for batch 0 and loss_mel * Add parameters from original VITS config * Fix config file * Fix imports and generate spec from audio * Fix incorrect dimensions * Progress update * Fix loss * Fix cuda thing * Fix monotonic align import * Fix typos in vits.py * Disable loss typecheck * Fix spectrogram lengths * Remove Precision 16 requirement * Address lgtm alerts * clean up unused code * Address lgtm alerts * Refactor audio_to_mel_torch method * Use NeMo FilterBank to get melspec Todo: set self.fb * Fix filterbank max frequency to match with original VITS * Fix filterbank features correct length * Address lgtm issues * Remove print statements * Remove stft_pad_amount Co-authored-by: martynwei <martyn.wei@gmail.com> Co-authored-by: Ryan Hong <66425733+rhong99@users.noreply.github.com> Co-authored-by: richa.ren@mail.utoronto.ca <richa.ren@mail.utoronto.ca> Co-authored-by: Jason <jasoli@nvidia.com> Signed-off-by: Jason <jasoli@nvidia.com> * make new commit Signed-off-by: Jason <jasoli@nvidia.com> * add copyright headers Signed-off-by: Jason <jasoli@nvidia.com> * style Signed-off-by: Jason <jasoli@nvidia.com> * rename README Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * fix style without vits_modules Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * add numba code, fix style and add todos Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> * small fix * fix some todos * added numba mas * added DDP sampler * specified versions * fixed for new librosa version * added feature loss * added IPA phonemizer * refactored IPA g2p * added vits losses * some ref * fix * added checkpointing * cp * cfg * merged some 1.8.0 fixes * plt fix * fix logging * fix checkpoint loading * refactored inference * fp32 run * update branch Signed-off-by: ericharper <complex451@gmail.com> * update package info Signed-off-by: ericharper <complex451@gmail.com> * new exp * update branch Signed-off-by: ericharper <complex451@gmail.com> * Restored tests previously disabled for 22.03 base (NVIDIA#4109) Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> * add augmentation to label models (NVIDIA#4113) * add augmentation to label models Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * duration fix Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * Call register_bert_model after assigning self.bert_model variable (NVIDIA#4116) Signed-off-by: Ramanathan Arunachalam <rarunachalam@nvidia.com> Co-authored-by: Ramanathan Arunachalam <rarunachalam@nvidia.com> * Tutorial on ITN with Thutmose tagger and small fixes (NVIDIA#4117) * 1. Add tutorial. 2. Move a function to fix import in tutorial. 3. Merge multiple spaces into one space in the final output Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * fixes for code review Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * Add tutorial to tutorials.rst Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> Co-authored-by: Alexandra Antonova <aleksandraa@nvidia.com> * cleaned up TN/ ITN doc (NVIDIA#4119) * cleaned up TN/ ITN doc Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix typo Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix image Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix image Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * Check implicit grad acc in GLUE dataset building (NVIDIA#4123) * Check implicit grad acc in GLUE dataset building Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix jenkins test for GLUE/XNLI Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * update the default (NVIDIA#4135) Signed-off-by: ekmb <ebakhturina@nvidia.com> * Draft: Fix restoring from checkpoint for case when `model.common_dataset_parameters.label_vocab_dir` is provided (NVIDIA#4136) * Fix restoring from checkpoint with label vocab dir Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Add tests for various ways to pass label ids to model Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Fix typo Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Fix typo Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Do not create tmp directory Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Fix parameter name Signed-off-by: PeganovAnton <peganoff2@mail.ru> * finish cherry-pick op Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Fix labels errors Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Remove duplicate stage Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Change target branch Signed-off-by: PeganovAnton <peganoff2@mail.ru> * fix typo (NVIDIA#4140) Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * Fix/punctuation avoid overwritting tmp files (NVIDIA#4144) * Add draft of fixing tmp files overwritting Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Remove accidental changes Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Remove accidental changes Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Use built-in tempfile library Signed-off-by: PeganovAnton <peganoff2@mail.ru> * Fix code style Signed-off-by: PeganovAnton <peganoff2@mail.ru> * bug_fix_diarization_manifest_creation (NVIDIA#4125) Signed-off-by: Yang Zhang <yangzhang@nvidia.com> Co-authored-by: Nithin Rao <nithinrao.koluguri@gmail.com> * fix doc (NVIDIA#4146) Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * Tacotron2 retrain (NVIDIA#4103) * fix yaml Signed-off-by: treacker <emshabalin@yandex.ru> * Fix for new TTSDataset class Signed-off-by: treacker <emshabalin@yandex.ru> * added wandb logging Signed-off-by: treacker <emshabalin@yandex.ru> * added wandb logging Signed-off-by: treacker <emshabalin@yandex.ru> * fix numpy version Signed-off-by: treacker <emshabalin@yandex.ru> * fix numpy version Signed-off-by: treacker <emshabalin@yandex.ru> * inference fix Signed-off-by: treacker <emshabalin@yandex.ru> * removed old code Signed-off-by: treacker <emshabalin@yandex.ru> * updated parser logic Signed-off-by: treacker <emshabalin@yandex.ru> * reverted version update Signed-off-by: treacker <emshabalin@yandex.ru> * refactored parser logic Signed-off-by: treacker <emshabalin@yandex.ru> * Updated Jenkinsfile Signed-off-by: treacker <emshabalin@yandex.ru> * Refactored tutorial for Tacotron2 Signed-off-by: treacker <emshabalin@yandex.ru> * Made backward compatibility Signed-off-by: treacker <emshabalin@yandex.ru> * Made backward compatibility Signed-off-by: treacker <emshabalin@yandex.ru> * Update Jenkinsfile Signed-off-by: treacker <emshabalin@yandex.ru> * Update tacotron.yaml Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> * cleaned up TN/ ITN doc (NVIDIA#4119) * cleaned up TN/ ITN doc Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix typo Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix image Signed-off-by: Yang Zhang <yangzhang@nvidia.com> * fix image Signed-off-by: Yang Zhang <yangzhang@nvidia.com> Signed-off-by: treacker <emshabalin@yandex.ru> * Check implicit grad acc in GLUE dataset building (NVIDIA#4123) * Check implicit grad acc in GLUE dataset building Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix jenkins test for GLUE/XNLI Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> * Fixed jenkins Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> * Refactoring Signed-off-by: treacker <emshabalin@yandex.ru> Co-authored-by: Yang Zhang <yzhang123@users.noreply.github.com> Co-authored-by: Sandeep Subramanian <sandeep.subramanian.1@umontreal.ca> * Multiprocess improvements (NVIDIA#4127) * initial commit Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * start fix Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * improve multiprocessing speed while creating speaker dataset Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * updated scp to filelist Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * WaveGlow input type fixes (NVIDIA#4151) Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> * notebooks' link, typo and import fix (NVIDIA#4158) * redo missing pr 4007 Signed-off-by: fayejf <fayejf07@gmail.com> * remove extremely unreliable links Signed-off-by: fayejf <fayejf07@gmail.com> * Thutmose tagger bug fixes (NVIDIA#4162) * add pretrained ngc model, small fixes Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * fix model location Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * fix model location Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * 1. fix typos. 2. write magic functions without space Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * add example of inference with pretrained model Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * changed model location to nemo Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * style fix Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> * fix space Signed-off-by: Alexandra Antonova <aleksandraa@nvidia.com> Co-authored-by: Alexandra Antonova <aleksandraa@nvidia.com> * update speaker docs (NVIDIA#4164) * update speaker docs Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * chunks -> segments Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * Khz -> kHz Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> * changed to vits g2p * refactoring * added cosineLR * Updated whitelist path * added vanilla torch grad scaler * Fixed lightning version * added warmup and wd * switched to cosineLR * refactored data classes for vits * some fixes * fixed import * changeg train loop * fixed scheduler bug * refactoring for exps * Refactored loss logic * Ref for exps * added coqui stuff * exps * bugfix * added side file * bugfix * reverted * fixed sampler behaviour * updated for ptl 1.7.2 * refactored dataloader func * some cleaning * reverted to vanilla loss * modified for pickling * added dataset class * fixed torch version * added autocast for fp training * removed coqui files * 'Fixed tokenizer' * Fix tokenizer * update branch Signed-off-by: ericharper <complex451@gmail.com> * Fix link to inference notebook (NVIDIA#5247) Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> * Update ASR scores table (NVIDIA#5254) Signed-off-by: smajumdar <titu1994@gmail.com> Signed-off-by: smajumdar <titu1994@gmail.com> * Fix links to speaker identification notebook (NVIDIA#5260) Signed-off-by: SeanNaren <snarenthiran@nvidia.com> Signed-off-by: SeanNaren <snarenthiran@nvidia.com> * Minor typo fixes in TTS tutorial (NVIDIA#5266) Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> * Pcla tutorial fixes (NVIDIA#5271) * Fixed typos Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Fixed cell type and tatoeba reference Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Fixed typo Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Fixed branch variable Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Fix bug into Dialogue tutorial (NVIDIA#5277) * Typo fix (NVIDIA#5288) Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Fix dialogue tutorial bug (NVIDIA#5297) * set add_pooling_layer=False for huggingface bert model * remove add_pooling_layer=False and set find_unused_parameters=True * set num_prompt_tokens to 0 for huggingface * small bugfix for r1.13.0 (NVIDIA#5310) * typo fix Signed-off-by: fayejf <fayejf07@gmail.com> * udpate transcribe Signed-off-by: fayejf <fayejf07@gmail.com> Signed-off-by: fayejf <fayejf07@gmail.com> * Add italian model checkpoints (NVIDIA#5316) Signed-off-by: Igor Gitman <igitman@nvidia.com> Signed-off-by: Igor Gitman <igitman@nvidia.com> * [STT] Add Ru ASR Conformer-CTC and Conformer-Transducer (NVIDIA#5340) * [STT] Add stt_ru_conformer_ctc_large Signed-off-by: Sasha Meister <117230141+ssh-meister@users.noreply.github.com> * [STT] Add stt_ru_conformer_transducer_large Add stt_ru_conformer_transducer_large Signed-off-by: Sasha Meister <117230141+ssh-meister@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Sasha Meister <117230141+ssh-meister@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Pcla tutorial fixes (NVIDIA#5313) * fixes Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * fixes Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * moved `create_text_and_labels` to token_classification_utils.py Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * a lot of refactoring * strict ptl version * strict ptl version * reverted plt version * Added base text2audio class * Fix issue with HF Model upload tutorial (NVIDIA#5359) * Add Gradio App to ASR Docs (NVIDIA#5270) Signed-off-by: smajumdar <titu1994@gmail.com> Signed-off-by: smajumdar <titu1994@gmail.com> (cherry picked from commit e4b6a38) * Fix issue with normalized config for dataset name Signed-off-by: smajumdar <titu1994@gmail.com> Signed-off-by: smajumdar <titu1994@gmail.com> * tutorial fixes (NVIDIA#5354) Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> * Add SDP documentation (NVIDIA#5274) * Add details to SDP README.md Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add docstring to WriteManifest processor Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add docstring to CreateInitialManifestMLS Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add ModifyManifestTextProcessor docstring Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add ASRInference docstring Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add base_processor docstrings Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Add minimal SDP docs page Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Update tools/speech_dataset_processor/README.md Co-authored-by: Igor Gitman <igitman@nvidia.com> Signed-off-by: Elena Rastorgueva <80532067+erastorgueva-nv@users.noreply.github.com> * Write simple README for SDP and move complex explanations to docs Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Remove incorrect type hints Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Make config example less confusing Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Fix typo Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Clarify that YAML file is config file in README Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Remove unused imports Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Remove SDP docs for now Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * Remove links to docs in SDP README Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> Signed-off-by: Elena Rastorgueva <80532067+erastorgueva-nv@users.noreply.github.com> Co-authored-by: Igor Gitman <igitman@nvidia.com> * [Bugfix] Added rm -f / wget- nc command in multispeaker sim notebook to r1.13.0 (NVIDIA#5375) * Fix minor error in notebook Signed-off-by: Taejin Park <tango4j@gmail.com> * changed branch name in tutorial notebook Signed-off-by: Taejin Park <tango4j@gmail.com> Signed-off-by: Taejin Park <tango4j@gmail.com> * Rename Speech Dataset Processor to Speech Data Processor (NVIDIA#5378) Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> * fix for num worker 0 causing issues in losses after 1 epoch (NVIDIA#5379) * Fixed bug in notebook (NVIDIA#5382) Signed-off-by: Virginia Adams <vadams@nvidia.com> Signed-off-by: Virginia Adams <vadams@nvidia.com> * Force MHA QKV onto fp32 (NVIDIA#5391) Signed-off-by: smajumdar <titu1994@gmail.com> Signed-off-by: smajumdar <titu1994@gmail.com> * Added scheduling variety * ref * Fix for prompt table restore error (NVIDIA#5393) * Fix for prompt table restore error Signed-off-by: Virginia Adams <vadams@nvidia.com> * Added more saftey checks Signed-off-by: Virginia Adams <vadams@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Added more condition checks Signed-off-by: Virginia Adams <vadams@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Virginia Adams <vadams@nvidia.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Fix args (NVIDIA#5410) Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * bugfix * import tests * Add temporary fix for CUDA issue in Dockerfile (NVIDIA#5421) Signed-off-by: Yu Yao <yuya@nvidia.com> Signed-off-by: Yu Yao <yuya@nvidia.com> * Megatron Export Update (NVIDIA#5343) * export update for Megatron + change ORT optimization Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * updated export_utils to use autocast instead of manually casting >:/ Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * removed dtype from LayerNorm Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * added comment Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * reverting changes on FloatCast Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * Cherry-picked changes from megatron-norm Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> * updated asr_model import to cast_utils Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * updated del onnx_model place Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * changed ort optimization to basic -> temp fix Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> Co-authored-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Boris Fomitchev <bfomitchev@nvidia.com> * disable pc test (NVIDIA#5426) Signed-off-by: ekmb <ebakhturina@nvidia.com> Signed-off-by: ekmb <ebakhturina@nvidia.com> * Fix GPT generation when using sentencepiece tokenizer (NVIDIA#5413) * Fix Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Co-authored-by: Yi Dong <yidong@nvidia.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> * Disable sync_batch_comm in validation_step for GPT (NVIDIA#5397) * disable sync_batch_comm in validation_step Signed-off-by: ericharper <complex451@gmail.com> * Read sync_batch_comm from config or default to False Signed-off-by: Markel Sanz Ausin <markelsanz14@gmail.com> * Update megatron_gpt_config to default sync_batch_comm to False to avoid CUDA error Signed-off-by: Markel Sanz Ausin <markelsanz14@gmail.com> * Empty Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Comment out test Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: ericharper <complex451@gmail.com> Signed-off-by: Markel Sanz Ausin <markelsanz14@gmail.com> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: Oleksii Kuchaiev <okuchaiev@nvidia.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> Co-authored-by: Markel Sanz Ausin <markelsanz14@gmail.com> Co-authored-by: Sandeep Subramanian <sandeep.subramanian.1@umontreal.ca> Co-authored-by: Oleksii Kuchaiev <okuchaiev@nvidia.com> * Revert "Add temporary fix for CUDA issue in Dockerfile (NVIDIA#5421)" (NVIDIA#5431) This reverts commit 0718b17. * Revert workaround for T5 that sets number of workers to 0 & sync_batch_comm=False (NVIDIA#5420) * Revert workers workaround Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix in config Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> * Fixed discrepancies * updated Jenkisfile * updated Jenkisfile * Cleaning * fixed the onnx bug in conformer for non-streaming models. (NVIDIA#5242) (NVIDIA#5446) Signed-off-by: Vahid <vnoroozi@nvidia.com> Signed-off-by: Vahid <vnoroozi@nvidia.com> Signed-off-by: Vladimir Bataev <vbataev@nvidia.com> Signed-off-by: Vahid <vnoroozi@nvidia.com> Signed-off-by: Vladimir Bataev <vbataev@nvidia.com> Co-authored-by: Vahid Noroozi <VahidooX@users.noreply.github.com> * Set sync_batch_comm in other places (NVIDIA#5448) Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Radtts 1.13 (NVIDIA#5451) * [TTS] Fixing RADTTS training - removing view buffer and fixing accuracy issue (NVIDIA#5358) * [TTS] add CI test for RADTTS training recipe. Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> Signed-off-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com> Co-authored-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> * Radtts 1.13 plus (NVIDIA#5457) * [TTS] Fixing RADTTS training - removing view buffer and fixing accuracy issue (NVIDIA#5358) * Fixing RADTTS training - removing view buffer and fixing accuracy issue * Fixes for Torchscript/Triton * Added autocast to radtts UT * using cuda() for training example Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> Signed-off-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com> Co-authored-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> * Add num layers check (NVIDIA#5470) Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Change to kwargs (NVIDIA#5475) Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Support for finetuning and finetuning inference with .ckpt files & batch size refactoring (NVIDIA#5339) (NVIDIA#5478) * Initial refactor Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Resolve config before passing to load_from_checkpoint Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fixes for model parallel and nemo restore Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fixes for eval Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Revert config changes Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Refactor Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix typo Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Remove comments Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Minor Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix validation reconfiguration Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Remove old comment Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fixes for test_ds Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * export_utils bugfix (NVIDIA#5480) * updated export_utils Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Export fixes for Riva (NVIDIA#5496) * Export fixes for Riva Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> * Cleaning up training_utils Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> * minor bug fix (NVIDIA#5521) Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: David Mosallanezhad <dmosallanezh@nvidia.com> * added set_start_method + function param bugfix (NVIDIA#5539) * added set_start_method + function param bugfix Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * upper bound torchmetrics Signed-off-by: ericharper <complex451@gmail.com> Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Signed-off-by: ericharper <complex451@gmail.com> Co-authored-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: ericharper <complex451@gmail.com> * remove notebook (NVIDIA#5548) Signed-off-by: ericharper <complex451@gmail.com> Signed-off-by: ericharper <complex451@gmail.com> * Remove broadcast (NVIDIA#5558) Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * cleaning * Fix all gather while writing to a file during T5 finetuning (NVIDIA#5561) * Gather from data parallel only instead of all ranks Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * Fix Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> * update readme Signed-off-by: ericharper <complex451@gmail.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * added copyright * fixed imports * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * cleaning * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fixed filesize check * last cleaning Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * updated cmudict path * fixed merge bug Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * warnings fix * fix warnings Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * storing * updated version Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * update Jenkinsfile versions Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * fixed issues Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * fixed more issues * more fixes Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * added experimental tag * Clarification updates Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * remove old cython code Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * remove old cython code Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * docstring fix Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * Enhancements Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * Enhancements Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * imports fix Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix typo Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * excessive comtutations fix Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * typecheck fix Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * Small refactoring * Small refactoring Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * reversed exp_manager params Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * Fixed call for new function signature Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by: Oktai Tatanov <oktai.tatanov@gmail.com> Signed-off-by: Jason <jasoli@nvidia.com> Signed-off-by: ericharper <complex451@gmail.com> Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com> Signed-off-by: nithinraok <nithinrao.koluguri@gmail.com> Signed-off-by: Yang Zhang <yangzhang@nvidia.com> Signed-off-by: MaximumEntropy <sandeep.subramanian.1@umontreal.ca> Signed-off-by: ekmb <ebakhturina@nvidia.com> Signed-off-by: PeganovAnton <peganoff2@mail.ru> Signed-off-by: Jocelyn Huang <jocelynh@nvidia.com> Signed-off-by: fayejf <fayejf07@gmail.com> Signed-off-by: smajumdar <titu1994@gmail.com> Signed-off-by: SeanNaren <snarenthiran@nvidia.com> Signed-off-by: Matvei Novikov <mattyson.so@gmail.com> Signed-off-by: Igor Gitman <igitman@nvidia.com> Signed-off-by: Sasha Meister <117230141+ssh-meister@users.noreply.github.com> Signed-off-by: Elena Rastorgueva <erastorgueva@nvidia.com> Signed-off-by: Elena Rastorgueva <80532067+erastorgueva-nv@users.noreply.github.com> Signed-off-by: Taejin Park <tango4j@gmail.com> Signed-off-by: Virginia Adams <vadams@nvidia.com> Signed-off-by: Yu Yao <yuya@nvidia.com> Signed-off-by: David Mosallanezhad <dmosallanezh@nvidia.com> Signed-off-by: Markel Sanz Ausin <markelsanz14@gmail.com> Signed-off-by: Oleksii Kuchaiev <okuchaiev@nvidia.com> Signed-off-by: Vahid <vnoroozi@nvidia.com> Signed-off-by: Vladimir Bataev <vbataev@nvidia.com> Signed-off-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com> Signed-off-by: Evgeniy Shabalin <baah1999@yandex.ru> Co-authored-by: jasonjjl1999 <jasonjjl1999@gmail.com> Co-authored-by: richa.ren@mail.utoronto.ca <richa.ren@mail.utoronto.ca> Co-authored-by: Oktai Tatanov <oktai.tatanov@gmail.com> Co-authored-by: jasonjjl1999 <43978361+jasonjjl1999@users.noreply.github.com> Co-authored-by: martynwei <martyn.wei@gmail.com> Co-authored-by: Ryan Hong <66425733+rhong99@users.noreply.github.com> Co-authored-by: Jason <jasoli@nvidia.com> Co-authored-by: ericharper <complex451@gmail.com> Co-authored-by: Boris Fomitchev <borisfom@users.noreply.github.com> Co-authored-by: Nithin Rao <nithinrao.koluguri@gmail.com> Co-authored-by: Ramanathan Arunachalam <ramanathan.arun@rutgers.edu> Co-authored-by: Ramanathan Arunachalam <rarunachalam@nvidia.com> Co-authored-by: bene-ges <61418381+bene-ges@users.noreply.github.com> Co-authored-by: Alexandra Antonova <aleksandraa@nvidia.com> Co-authored-by: Yang Zhang <yzhang123@users.noreply.github.com> Co-authored-by: Sandeep Subramanian <sandeep.subramanian.1@umontreal.ca> Co-authored-by: Evelina <10428420+ekmb@users.noreply.github.com> Co-authored-by: PeganovAnton <peganoff2@mail.ru> Co-authored-by: Jocelyn <jocelynh@nvidia.com> Co-authored-by: fayejf <36722593+fayejf@users.noreply.github.com> Co-authored-by: Somshubra Majumdar <titu1994@gmail.com> Co-authored-by: Sean Naren <snarenthiran@nvidia.com> Co-authored-by: Matvei Novikov <mattyson.so@gmail.com> Co-authored-by: Zhilin Wang <wangzhilin12061996@hotmail.com> Co-authored-by: Igor Gitman <igitman@nvidia.com> Co-authored-by: Sasha Meister <117230141+ssh-meister@users.noreply.github.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Elena Rastorgueva <80532067+erastorgueva-nv@users.noreply.github.com> Co-authored-by: Taejin Park <tango4j@gmail.com> Co-authored-by: Adi Renduchintala <108822655+arendu@users.noreply.github.com> Co-authored-by: Virginia Adams <78445382+vadam5@users.noreply.github.com> Co-authored-by: yaoyu-33 <54727607+yaoyu-33@users.noreply.github.com> Co-authored-by: David <amosalla@asu.edu> Co-authored-by: David Mosallanezhad <dmosallanezh@nvidia.com> Co-authored-by: Boris Fomitchev <bfomitchev@nvidia.com> Co-authored-by: Yi Dong <yidong@nvidia.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@users.noreply.github.com> Co-authored-by: Markel Sanz Ausin <markelsanz14@gmail.com> Co-authored-by: Oleksii Kuchaiev <okuchaiev@nvidia.com> Co-authored-by: Vladimir Bataev <vbataev@nvidia.com> Co-authored-by: Vahid Noroozi <VahidooX@users.noreply.github.com> Co-authored-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com>
- Loading branch information
43 people
committed
Mar 24, 2023
1 parent
d9b8f38
commit 80e92a4