Comments (5)
Yeah, this PTL MNIST example on a gpu with max_nb_epochs=1 will finish in a few seconds.
To watch it live you need something that takes longer to train.
Remove max_nb_epochs or set min_nb_epochs=100 (you’ll overfit fyi) to force it to train for much longer so u can watch the live logs.
from lightning.
Happy to reopen this issue if this doesn’t fix it. Enjoy the lightning!
from lightning.
On my local machine the same code updates the tensorboard file as expected every 100 samples.
Is it possible that there are any handles on the file that are open during training, but closed after? I remember a similar problem, where Google Colab Notebooks updated files only if all handles were closed.
from lightning.
yeah, i think PyTorch SummaryWriter keeps a file open for saving.
But i suspect the issue is with how Colab handles open files. PTL doesn’t keep any files open, it defaults to whatever the SummaryWriter behavior is.
Glad to hear it’s fine on your local though.
If you find out what the issue is on colab, def follow up here to see if there’s anything we can do on our end
from lightning.
@alex-obi update to latest test-tube version and try again?
pt 1.2.0
test tube 0.6.9
lightning 0.4.3
from lightning.
Related Issues (20)
- Mixing the order of `--config` and `fit` in LightningCLI can cause confusion HOT 7
- Support TorchEval HOT 2
- Extending Argument Linking for more complicate relationship HOT 2
- Training stuck when running on Slurm with multiprocessing HOT 1
- PyTorch Lightning FSDP takes more memory than PyTorch FSDP HOT 5
- Incompatible LightningModule.to_onnx type-hint HOT 1
- LightningCLI, WandbLogger and save_hyperparameters() inconsistencies
- `log_dir` contains both forward and backward slashes as path separator when using remote file location as `default_root_dir` on windows
- ValueError: dictionary update sequence element #0 has length 1; 2 is required
- Expected all tensors to be on the same device... using quantized model with deepspeed zero3
- Huggingface model quantization has odd behavior with BitsandbyesPrecisionPlugin HOT 3
- `TensorBoardLogger` fails with remote FS (azure)
- Pydantic AttributeError when using LightningDataModule
- Loading from a checkpoint does not work properly in distributed training
- Optimizer step based validation trigger HOT 1
- Log `TensorBoard` histograms
- PTL 2.2 specifically causes torchscript errors when loaded in any environment not containing PTL 2.2
- When calling trainer.test() train_dataloader is also validated, which makes no sense HOT 2
- Support `ThunderModule` models HOT 2
- Calculated loss differs from logged loss in training_step (even if seed_everything, deterministic set to true and shuffle to false) HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from lightning.