Discussions

• 1 You must be logged in to vote

  💭

  HuggingFace Transformers

  chris-boson asked Mar 10, 2021 in code help: NLP / ASR / TTS · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🤖

  How to use DDP + FP16 without fit?

  turian asked Mar 10, 2021 in DDP / multi-GPU / multi-node · Unanswered

  

  

  2
• 1 You must be logged in to vote

  ⚡

  How to get global step from checkpoint?

  timothybrooks asked Mar 10, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  1
• 1 You must be logged in to vote

  ⚡

  Pretrain some sections of a model, then initialize those parts when training a full model

  stanleyshly asked Mar 10, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  1
• 1 You must be logged in to vote

  ⚡

  Scheduler.step() only called on the end of validation

  richarddwang asked Mar 10, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  2
• 2 You must be logged in to vote

  ⚡

  import pytorch_lightning as pl does not work on colab

  Bajo1994 asked Mar 9, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  

  

  2
• 1 You must be logged in to vote

  ⚡

  how to print loss every n steps without progress bar

  sun-peach asked Mar 10, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  1
• 1 You must be logged in to vote

  ⚡

  How set number of epochs

  Vichoko asked Nov 8, 2019 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  

  

  4
• 1 You must be logged in to vote

  ⚡

  changing val_check_interval from 0.01 to 0.005 changes number of steps in an epoch

  AyushP123 asked Mar 9, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  1
• 1 You must be logged in to vote

  🤖

  Predict on GPU and calculate metrics in CPU, DDP Mode

  frapercan asked Mar 9, 2021 in DDP / multi-GPU / multi-node · Unanswered

  

  

  0
• 2 You must be logged in to vote

  🤖

  DDP number of training iterations independent of #GPUs?

  timothybrooks asked Mar 3, 2021 in DDP / multi-GPU / multi-node · Unanswered

  

  

  5
• 1 You must be logged in to vote

  ⚡

  Should Lightning have separate setup and teardown hooks for each stage?

  ananthsub asked Mar 7, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  

  

  4
• 1 You must be logged in to vote

  ⚡

  Access datamodule in custom callbacks

  vishal-wiai asked Mar 8, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  1
• 1 You must be logged in to vote

  ⚡

  How to correctly apply metrics API in binary use case

  kapsner asked Mar 5, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  7
• 1 You must be logged in to vote

  🤖

  How to sequentially call fit() and test() in DDP

  kapsner asked Mar 5, 2021 in DDP / multi-GPU / multi-node · Answered

  

  

  2
• 2 You must be logged in to vote

  ⚡

  Lightning isn`t automatically moving a tensor to certain device

  accelerator: tpu Tensor Processing Unit data handling Generic data-related topic

  StephennFernandes asked Mar 7, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  3
• 1 You must be logged in to vote

  ⚡

  Continuing training resets logger epoch

  wittenator asked Mar 7, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  0
• 1 You must be logged in to vote

  ⚡

  boring model for DDP

  hkmztrk asked Mar 5, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  

  

  6
• 1 You must be logged in to vote

  🤖

  Deterministic DataLoader on DDP reads same data on all subprocesses

  mrmotallebi asked Mar 4, 2021 in DDP / multi-GPU / multi-node · Answered

  

  

  1
• 1 You must be logged in to vote

  ⚡

  val_check_interval and check_val_every_n_epoch appear redundant?

  turian asked Feb 28, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  10
• 1 You must be logged in to vote

  ⚡

  How to have model defined in another file?

  stanleyshly asked Mar 6, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  0
• 1 You must be logged in to vote

  💬

  Unable to use Trainer.predict: 'Trainer' object has no attribute 'predict_dataloaders'

  its-dron started Mar 5, 2021 in General

  

  

  1
• 1 You must be logged in to vote

  ⚡

  How to save a checkpoint every n steps and overwrite the previous saved checkpoints?

  stanleyshly asked Mar 3, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  3
• 2 You must be logged in to vote

  ⚡

  Different metric result in callback and checkpoint?

  rudaoshi asked Mar 5, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Unanswered

  

  

  0
• 2 You must be logged in to vote

  ⚡

  Access and change models optimizer after setup

  Haydnspass asked Feb 22, 2021 in Lightning Trainer API: Trainer, LightningModule, LightningDataModule · Answered

  

  

  3