Nantes Université

Skip to content
Extraits de code Groupes Projets
Valider a0f90b96 rédigé par Léonard Seydoux's avatar Léonard Seydoux
Parcourir les fichiers

Test push

parent ba95e0f7
Aucune branche associée trouvée
Aucune étiquette associée trouvée
Aucune requête de fusion associée trouvée
NuTS - Machine Learning workshop TP
===================================
This TP is based on [L. Seydoux SPIN-SC3 course available on GitHub](https://github.com/spin-itn/SPIN-SC3).
This TP is based on [Leonard Seydoux's SPIN-SC3 course available on GitHub](https://github.com/spin-itn/SPIN-SC3).
It was adapted for the NuTS workshop that will be held in Lyon in June 2023.
The workshop will be held in a dedicated JupyterHub environment pre-configured for all the users.
......@@ -17,6 +17,7 @@ Your file manager is located on the left and your notebook on the right.
Local testing
-------------
If you want to test these notebooks locally, you will need to have `conda` or `mamba` installed locally on your computer.
Then clone this repo, install the required dependencies and launch a JupyterLab environment where you can run the notebooks:
......@@ -32,12 +33,15 @@ conda activate nuts-workshop
# Start Jupyter
SEISBENCH_CACHE_ROOT="/absolute/path/to/seisbench/data/" jupyter lab
```
__Check list:__
- We encore you to download the Seisbench data in advance:
- either the large dataset [s3.glicid.fr/nuts/seisbench-data.zip](https://s3.glicid.fr/nuts/seisbench-data.zip) (6 GB compressed, 29 GB uncompressed)
- or a smaller one [s3.glicid.fr/nuts/seisbench-data-small.zip](https://s3.glicid.fr/nuts/seisbench-data-small.zip) (1.5 GB compressed, 5 GB uncompressed)
- either the large dataset [s3.glicid.fr/nuts/seisbench-data.zip](https://s3.glicid.fr/nuts/seisbench-data.zip) (6 GB compressed, 29 GB uncompressed)
- or a smaller one [s3.glicid.fr/nuts/seisbench-data-small.zip](https://s3.glicid.fr/nuts/seisbench-data-small.zip) (1.5 GB compressed, 5 GB uncompressed)
Then unzip the archive and fill the `SEISBENCH_CACHE_ROOT` environment variable. You can check that the data are available like this:
```python
import os
......@@ -54,6 +58,7 @@ data.metadata.head()
```
- Use of a dedicated GPU is not required by highly encouraged. You can check if you GPU is correctly detected by `pytorch` like this:
```python
import torch
......
0% Chargement en cours ou .
You are about to add 0 people to the discussion. Proceed with caution.
Terminez d'abord l'édition de ce message.
Veuillez vous inscrire ou vous pour commenter