Lightning-AI/lightning issues and pull requests

#4462 - Is there a way to supress all printing and warnings in the Trainer?

Issue - State: closed - Opened by erap129 almost 4 years ago - 3 comments
Labels: question

#4451 - Rename conflicting test directories

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 1 comment
Labels: feature, ready, ci

#4450 - Data loading hangs before first validation step

Issue - State: closed - Opened by jonashaag almost 4 years ago - 27 comments
Labels: help wanted, won't fix, waiting on author

#4435 - Fix AttributeError with num_processes>1 but no ddp_cpu

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 1 comment

#4420 - NCCL error using DDP and PyTorch 1.7

Issue - State: closed - Opened by ohmeow almost 4 years ago - 54 comments
Labels: bug, help wanted, priority: 0, distributed, 3rd party

#4419 - Fix CSV logger warning

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 1 comment
Labels: bug, ready

#4383 - Add "monitor" to saved ModelCheckpoints

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 2 comments
Labels: feature, ready, checkpointing

#4109 - Improve skipping steps tests

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 8 comments
Labels: ci

#4108 - Mention skipping steps in docs

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 2 comments
Labels: docs

#4045 - continue training from checkpoint seems broken (high loss values), while reasonable with .eval()

Issue - State: closed - Opened by yairkit almost 4 years ago - 17 comments
Labels: bug, help wanted, priority: 0

#4045 - continue training from checkpoint seems broken (high loss values), while reasonable with .eval()

Issue - State: closed - Opened by yairkit almost 4 years ago - 17 comments
Labels: bug, help wanted, priority: 0

#3796 - Fix tbptt_reduce_fx when non-floating tensors are logged

Pull Request - State: closed - Opened by carmocca almost 4 years ago - 3 comments
Labels: bug

#3644 - Split GPUStatsMonitor function

Pull Request - State: closed - Opened by carmocca about 4 years ago - 3 comments
Labels: feature, ready

#3633 - Allow ModelCheckpoint monitor to be None

Pull Request - State: closed - Opened by carmocca about 4 years ago - 2 comments
Labels: feature, ready

#3630 - Fix ModelCheckpoint period

Pull Request - State: closed - Opened by carmocca about 4 years ago - 28 comments
Labels: bug, ready

#3627 - Allow monitoring tensors logged via TrainResult

Pull Request - State: closed - Opened by carmocca about 4 years ago - 2 comments

#3625 - Log validation metrics before training

Issue - State: closed - Opened by bartonp2 about 4 years ago - 8 comments
Labels: question, won't fix

#3621 - Support multiple checkpoint callbacks

Pull Request - State: closed - Opened by carmocca about 4 years ago - 9 comments

#3595 - Update gitignore

Pull Request - State: closed - Opened by carmocca about 4 years ago - 1 comment
Labels: feature, ready

#3594 - testing bug from #3163

Pull Request - State: closed - Opened by carmocca about 4 years ago - 3 comments
Labels: bug, ci

#3588 - Fix incorrect "Saving latest checkpoint" warning

Pull Request - State: closed - Opened by carmocca about 4 years ago - 5 comments
Labels: bug, ready, checkpointing

#3587 - Clean up optimizer code

Pull Request - State: closed - Opened by carmocca about 4 years ago - 6 comments
Labels: feature, ready, refactor

#3586 - Add strict option to lr_scheduler dict

Pull Request - State: closed - Opened by carmocca about 4 years ago - 7 comments
Labels: feature, ready

#3566 - Support skipping training_step

Pull Request - State: closed - Opened by carmocca about 4 years ago - 31 comments
Labels: feature

#3565 - Replace log_gpu_memory with gpu_stats_callback

Pull Request - State: closed - Opened by carmocca about 4 years ago - 4 comments
Labels: feature, won't fix, design

#3528 - Fix log debug call

Pull Request - State: closed - Opened by carmocca about 4 years ago - 1 comment
Labels: bug, ready

#3485 - ModelCheckpoint every validation cycle instead of epoch?

Issue - State: closed - Opened by dvirginz about 4 years ago - 3 comments
Labels: question

#3235 - Parse Union[bool, str] arguments

Pull Request - State: closed - Opened by carmocca about 4 years ago - 4 comments
Labels: bug

#3228 - Log epoch as step when on_epoch=True and on_step=False

Issue - State: closed - Opened by ToucheSir about 4 years ago - 31 comments
Labels: feature, help wanted

#3173 - Set pep8speaks' max-line-length to 120 (same as black)

Pull Request - State: closed - Opened by carmocca about 4 years ago - 1 comment
Labels: ready, ci

#3163 - Fix ModelCheckpoints name formatting

Pull Request - State: closed - Opened by carmocca about 4 years ago - 28 comments
Labels: bug, feature, ready

#3148 - Extend DataLoader with transform arguments to keep that logic in the DataModule

Issue - State: closed - Opened by NumesSanguis about 4 years ago - 9 comments
Labels: feature, help wanted, discussion, data handling