Working from home
Highlights
- 53 discussions answered
Block or Report
Block or report rohitgr7
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePinned
-
The lightweight PyTorch wrapper for high-performance AI research. Scale your models, not the boilerplate.
-
opencv-filters Public
Implementation of image filters with opencv. Contains both image and video filters.
-
1,180 contributions in the last year
Less
More
Activity overview
Contributed to
PyTorchLightning/pytorch-lightning,
PyTorchLightning/metrics,
pytorch/pytorch
and 5 other
repositories
Contribution activity
January 2022
Created 28 commits in 3 repositories
Created 1 repository
- rohitgr7/pytorch-lightning Python
Created a pull request in PyTorchLightning/pytorch-lightning that received 4 comments
Update debugging doc
What does this PR do? Thanks, @ananthsub. Missed it. BTW anything else do you think is worth adding? #11390 (review) Changes: Added detect anomaly…
+103
−47
•
4
comments
Opened 15 other pull requests in 1 repository
PyTorchLightning/pytorch-lightning
9
open
6
merged
- added warning for distributedsampler in case of evaluation
-
Update
load_from_checkpointdocstrings - Update Model Parallel doc
- Add LightningModule hooks chart
- Update utilities API references
- Update profiler doc (1/n)
- Update profiler doc (2/n)
- Update inference doc
- improve simple profiler output
- Pin sphinx-autodoc-typehints with <v1.15
- Update PT to PL conversion doc
- Refine style guide
- Refine remote fs doc
- Refine debugging doc
- Update API references in doc
Reviewed 68 pull requests in 3 repositories
PyTorchLightning/pytorch-lightning
58 pull requests
- Update profiler doc (2/n)
-
Set
Loop.restartingrecursively - Refine style guide
- Update speed docs
-
Update
tests/utilities/*.pyto usedevicesinstead ofgpusoripus - Add XLA Profiler section to docs
- Fix compatibility with old checkpoints and fault-tolerance enabled
- Update quick start docs
- Add typing to accelerators/gpu.py
-
Update
tests/core/*.pyto usedevicesinstead ofgpusoripus - Update PT to PL conversion doc
- Refine remote fs doc
- Add Lightning Ecosystem to docs
- Update inference doc
- improve simple profiler output
-
Fix
torch.distributed._*import statements in tests - Update evaluation docs
- update tutorials
- Avoid in-place ops during logging result updates
-
Update
tests/callbacks/*.pyto usedevicesinstead ofgpusoripus -
Add typing to
TQDMProgressBar -
Add
LightningModule.lr_scheduler_step - Update mergify team name
- Update child modules docs
- Move progress bar disabling out of the Trainer
- Some pull request reviews not shown.
PyTorchLightning/metrics
9 pull requests
PyTorchLightning/ecosystem-ci
1 pull request
Opened 2 issues in 1 repository
PyTorchLightning/pytorch-lightning
2
open
Answered 11 discussions in 1 repository
PyTorchLightning/pytorch-lightning
PyTorchLightning/pytorch-lightning
- Confusion in training_step_end() API
- trainer.test( ) not working
- PyTorch Lightning Optimizer_Step() prevents training_step() from running
- PyTorch Lightning Optimizer_Step() prevents training_step() from running
- what's the difference between `load_from_checkpoint ` and `resume_from_checkpoint`
- self.manual_backward() vs. loss.backward() when optimizing manually
- Why is on_before_optimizer_step incompatible with accumulate_grad_batches ?
- Why is on_before_optimizer_step incompatible with accumulate_grad_batches ?
- ddp: how to combine multi-gpus outputs like "training_step_end" which is only used in dp/ddp2?
- Where to transform and inverse-transform
- Where to transform and inverse-transform