https://github.com/neuro-ml/thunder
đŠī¸ The Deep Learning framework based on Lightning
https://github.com/neuro-ml/thunder
deep-learning experiments machine-learning python pytorch pytorch-lightning
Last synced: 10 months ago
JSON representation
đŠī¸ The Deep Learning framework based on Lightning
- Host: GitHub
- URL: https://github.com/neuro-ml/thunder
- Owner: neuro-ml
- License: mit
- Created: 2023-05-31T21:43:18.000Z (almost 3 years ago)
- Default Branch: master
- Last Pushed: 2025-04-26T10:03:56.000Z (11 months ago)
- Last Synced: 2025-05-28T03:39:57.458Z (10 months ago)
- Topics: deep-learning, experiments, machine-learning, python, pytorch, pytorch-lightning
- Language: Python
- Homepage: https://neuro-ml.github.io/thunder/
- Size: 2.6 MB
- Stars: 10
- Watchers: 2
- Forks: 1
- Open Issues: 30
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
[](https://neuro-ml.github.io/thunder/)

[](https://codecov.io/gh/neuro-ml/thunder)
[](https://pypi.org/project/thunder/)
> _You saw the lightning. Now it's time to hear the thunder_ đŠī¸
# Thunder đŠī¸
The Deep Learning framework based on [Lightning](https://github.com/Lightning-AI/lightning).
## Install
```bash
pip install thunder
```
:warning:
> Currently thunder is not published on pypi. Install it via git clone.
## Start experimenting
It's as simple as 1, 2, 3:
1. Create a config (e.g. `base.config`):
```python
from myproject import MyDataset, MyModule
from lightning import Trainer
from torch.utils.data import DataLoader
# these 3 fields are required
train_data = DataLoader(MyDataset())
module = MyModule()
trainer = Trainer()
```
2. Build the experiment:
```shell
thunder build base.config /path/to/some/folder
```
3. Run it
```shell
thunder run /path/to/some/folder
```
Also, 2 and 3 can be combined into a single command:
```shell
thunder build-run base.config /path/to/some/folder
```
## More advanced stuff
See [our docs](https://neuro-ml.github.io/thunder/) for a full list of neat things `thunder`đŠī¸ can do for you