carmocca Follow

carmocca Follow

🗿

¡Hola!

Carlos Mocholí carmocca

🗿

¡Hola!

Follow

Research Engineer

42 followers · 16 following · 48

@PyTorchLightning
Valencia, Spain
https://www.linkedin.com/in/carlosmocholi

Highlights

35 discussions answered

Organizations

Block or Report

Block or report carmocca

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse

Pinned

jpuigcerver/PyLaia

A deep learning toolkit specialized for handwritten document analysis

Python 78 20
PyTorchLightning/pytorch-lightning

The lightweight PyTorch wrapper for high-performance AI research. Scale your models, not the boilerplate.

Python 14.6k 1.7k
PyLaia-examples

A set of experiments using PyLaia on different datasets

Shell 3
PyTorchLightning/lightning-flash

Collection of tasks for fast prototyping, baselining, finetuning and solving problems with deep learning.

Python 868 84
UVA

UVA programming challenges

Python 2 3
PyTorchLightning/lightning-transformers

Flexible interface for high-performance research using SOTA Transformers leveraging Pytorch Lightning, Transformers, and Hydra.

Python 309 26

2,358 contributions in the last year

Learn how we count contributions

Less More

@PyTorchLightning

Activity overview

Contributed to PyTorchLightning/pytorch-lightning, jpuigcerver/PyLaia, PyTorchLightning/lightning-transformers and 5 other repositories

Contribution activity

July 2021

Created 78 commits in 2 repositories

Created a pull request in PyTorchLightning/pytorch-lightning that received 8 comments

`every_n_val_epochs` -> `every_n_epochs`

What does this PR do? Deprecate every_n_val_epochs in favor of every_n_epochs The flag is used in the on_validation_end hook, but we will also want…

+91 −72 • 8 comments

Opened 42 other pull requests in 2 repositories

PyTorchLightning/pytorch-lightning

37 merged 4 open

Add pyupgrade to pre-commit Jul 26
Fix profiler test on Windows minimal Jul 26
Connect the model to the training type plugin at the start of run Jul 23
Avoid partial for apply to collection Jul 23
Update issue and PR templates Jul 23
Fix DeepSpeed lr scheduler logic Jul 23
Remove torch >= 1.6 checks Jul 22
Support DataLoaders with missing arguments in replace_sampler Jul 22
Always use trainer.call_hook Jul 20
Replace iteration_count and other index attributes in the loops with progress dataclasses Jul 19
Do not reset Loops total counters Jul 19
Unblock GPU CI Jul 18
Delete legacy DataLoader processing utility Jul 16
Move plateau schedulers epoch update to the training epoch loop Jul 14
Mark evaluation epoch loops attributes as protected Jul 14
Refactor log_dir usage in the CLI Jul 14
Remove pep8speaks Jul 13
Add pydocstyle to pre-commit Jul 13
Add ModelCheckpoint(save_on_train_epoch_end) Jul 13
Update to Mypy>0.9 Jul 12
Clean code formatting CI job Jul 12
Fix broadcast for Windows minimal Jul 7
Refactor plugins backward Jul 7
Fix self.optimizers() not returning a single LightningOptimizer Jul 7
Unpin Pillow after the 8.3.1 release Jul 7
Some pull requests not shown.

gridai/grid-docs

1 merged

Update CLI hint about feature parity Jul 21

Reviewed 120 pull requests in 2 repositories

PyTorchLightning/pytorch-lightning 119 pull requests

Connect the model to the training type plugin at the start of run Jul 26
black: magic trailing comma Jul 26
docs: explain how Lightning uses closures for automatic optimization Jul 26
v1.4.0rc2 Jul 26
Replace iteration_count and other index attributes in the loops with progress dataclasses Jul 26
Legacy: simple classif training Jul 26
docs: clarify closure usage in gan example Jul 23
Raise exception for ddp_cpu not supported for TPUs Jul 23
Support DataLoaders with missing arguments in replace_sampler Jul 23
[Typo] update some out-dated links from pytorch clip_grad_value_ Jul 23
Update issue and PR templates Jul 23
checkpoint also QAT for resuming Jul 23
Fix DeepSpeed lr scheduler logic Jul 23
fix CI for PT 1.10 Jul 23
fix restoring finetune callbacks after accelerator setup on training resume Jul 22
docs: fix return type description of Trainer.validate/test Jul 22
Add ddp_*_find_unused_parameters_false to Plugins Registry. Jul 22
Deprecate LightningModule.summarize() in favor of pytorch_lightning.utilities.model_summary.summarize() Jul 21
Add support for functions to be parsed by the Lightning CLI in addition to Types Jul 21
Refactor plugins backward Jul 21
Quant as optional step Jul 21
fix: Enable manual optimization for TPUs Jul 21
Always use trainer.call_hook Jul 21
[bugfix] Reduce memory leaks Jul 20
v1.4.0rc1 & chlog Jul 20
Some pull request reviews not shown.

PyTorchLightning/metrics 1 pull request

Update docs for new template Jul 1

Opened 1 issue in 1 repository

PyTorchLightning/pytorch-lightning

1 open

Re-design call_hook interface Jul 21

Answered 10 discussions in 1 repository

PyTorchLightning/pytorch-lightning

8 contributions in private repositories Jul 21 – Jul 22