{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"execution": {
"iopub.execute_input": "2021-01-08T00:45:39.077620Z",
"iopub.status.busy": "2021-01-08T00:45:39.077079Z",
"iopub.status.idle": "2021-01-08T00:45:39.085072Z",
"shell.execute_reply": "2021-01-08T00:45:39.084633Z"
},
"jupyter": {
"source_hidden": true
}
},
"outputs": [
{
"data": {
"text/markdown": [
"<a name=\"top\"></a>\n",
"\n",
"[<img width=\"600px\" src=\"fidle/img/00-Fidle-titre-01.svg\"></img>](#top)\n",
"\n",
"<!-- --------------------------------------------------- -->\n",
"<!-- To correctly view this README under Jupyter Lab -->\n",
"<!-- Open the notebook: README.ipynb! -->\n",
"<!-- --------------------------------------------------- -->\n",
"\n",
"\n",
"## A propos\n",
"\n",
"This repository contains all the documents and links of the **Fidle Training** . \n",
"Fidle (for Formation Introduction au Deep Learning) is a 2-day training session \n",
"co-organized by the Formation Permanente CNRS and the SARI and DEVLOG networks. \n",
"\n",
"The objectives of this training are :\n",
" - Understanding the **bases of Deep Learning** neural networks\n",
" - Develop a **first experience** through simple and representative examples\n",
" - Understanding **Tensorflow/Keras** and **Jupyter lab** technologies\n",
" - Apprehend the **academic computing environments** Tier-2 or Tier-1 with powerfull GPU\n",
"\n",
"For more information, you can contact us at : \n",
"[<img width=\"200px\" style=\"vertical-align:middle\" src=\"fidle/img/00-Mail_contact.svg\"></img>](#top) \n",
"Current Version : <!-- VERSION_BEGIN -->\n",
"1.2b1 DEV\n",
"<!-- VERSION_END -->\n",
"\n",
"\n",
"## Course materials\n",
"\n",
"| | | |\n",
"|:--:|:--:|:--:|\n",
"| **[<img width=\"50px\" src=\"fidle/img/00-Fidle-pdf.svg\"></img><br>Course slides](https://cloud.univ-grenoble-alpes.fr/index.php/s/wxCztjYBbQ6zwd6)**<br>The course in pdf format<br>(12 Mo)| **[<img width=\"50px\" src=\"fidle/img/00-Notebooks.svg\"></img><br>Notebooks](https://gricad-gitlab.univ-grenoble-alpes.fr/talks/fidle/-/archive/master/fidle-master.zip)**<br> Get a Zip or clone this repository <br>(10 Mo)| **[<img width=\"50px\" src=\"fidle/img/00-Datasets-tar.svg\"></img><br>Datasets](https://cloud.univ-grenoble-alpes.fr/index.php/s/wxCztjYBbQ6zwd6)**<br>All the needed datasets<br>(1.2 Go)|\n",
"\n",
"Have a look about **[How to get and install](https://gricad-gitlab.univ-grenoble-alpes.fr/talks/fidle/-/wikis/Install-Fidle)** these notebooks and datasets.\n",
"\n",
"\n",
"## Jupyter notebooks\n",
"\n",
"<!-- INDEX_BEGIN -->\n",
"<style>\n",
"\n",
".fid_line{\n",
" padding-top: 10px\n",
"}\n",
"\n",
".fid_id { \n",
" font-size:1.em;\n",
" color:black;\n",
" font-weight: bold; \n",
" padding:0px;\n",
" margin-left: 20px;\n",
" display: inline-block;\n",
" width: 60px;\n",
" }\n",
"\n",
".fid_desc { \n",
" font-size:1.em;\n",
" padding:0px;\n",
" margin-left: 85px;\n",
" display: inline-block;\n",
" width: 600px;\n",
" }\n",
"\n",
"\n",
"\n",
"div.fid_section { \n",
" font-size:1.2em;\n",
" color:black;\n",
" margin-left: 0px;\n",
" margin-top: 12px;\n",
" margin-bottom:8px;\n",
" border-bottom: solid;\n",
" border-block-width: 1px;\n",
" border-block-color: #dadada;\n",
" width: 700px;\n",
" }\n",
"\n",
"</style>\n",
"<div class=\"fid_section\">Linear and logistic regression</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"LinearReg/01-Linear-Regression.ipynb\">LINR1</a>\n",
" </span> <a href=\"LinearReg/01-Linear-Regression.ipynb\">Linear regression with direct resolution</a><br>\n",
" <span class=\"fid_desc\">Direct determination of linear regression </span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"LinearReg/02-Gradient-descent.ipynb\">GRAD1</a>\n",
" </span> <a href=\"LinearReg/02-Gradient-descent.ipynb\">Linear regression with gradient descent</a><br>\n",
" <span class=\"fid_desc\">An example of gradient descent in the simple case of a linear regression.</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"LinearReg/03-Polynomial-Regression.ipynb\">POLR1</a>\n",
" </span> <a href=\"LinearReg/03-Polynomial-Regression.ipynb\">Complexity Syndrome</a><br>\n",
" <span class=\"fid_desc\">Illustration of the problem of complexity with the polynomial regression</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"LinearReg/04-Logistic-Regression.ipynb\">LOGR1</a>\n",
" </span> <a href=\"LinearReg/04-Logistic-Regression.ipynb\">Logistic regression, with sklearn</a><br>\n",
" <span class=\"fid_desc\">Logistic Regression using Sklearn</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Perceptron Model 1957</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"IRIS/01-Simple-Perceptron.ipynb\">PER57</a>\n",
" </span> <a href=\"IRIS/01-Simple-Perceptron.ipynb\">Perceptron Model 1957</a><br>\n",
" <span class=\"fid_desc\">A simple perceptron, with the IRIS dataset.</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Basic regression using DNN</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"BHPD/01-DNN-Regression.ipynb\">BHPD1</a>\n",
" </span> <a href=\"BHPD/01-DNN-Regression.ipynb\">Regression with a Dense Network (DNN)</a><br>\n",
" <span class=\"fid_desc\">A Simple regression with a Dense Neural Network (DNN) - BHPD dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"BHPD/02-DNN-Regression-Premium.ipynb\">BHPD2</a>\n",
" </span> <a href=\"BHPD/02-DNN-Regression-Premium.ipynb\">Regression with a Dense Network (DNN) - Advanced code</a><br>\n",
" <span class=\"fid_desc\">More advanced example of DNN network code - BHPD dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Basic classification using a DNN</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"MNIST/01-DNN-MNIST.ipynb\">MNIST1</a>\n",
" </span> <a href=\"MNIST/01-DNN-MNIST.ipynb\">Simple classification with DNN</a><br>\n",
" <span class=\"fid_desc\">Example of classification with a fully connected neural network</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Images classification with Convolutional Neural Networks (CNN)</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/01-Preparation-of-data.ipynb\">GTSRB1</a>\n",
" </span> <a href=\"GTSRB/01-Preparation-of-data.ipynb\">CNN with GTSRB dataset - Data analysis and preparation</a><br>\n",
" <span class=\"fid_desc\">Episode 1 : Data analysis and creation of a usable dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/02-First-convolutions.ipynb\">GTSRB2</a>\n",
" </span> <a href=\"GTSRB/02-First-convolutions.ipynb\">CNN with GTSRB dataset - First convolutions</a><br>\n",
" <span class=\"fid_desc\">Episode 2 : First convolutions and first results</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/03-Tracking-and-visualizing.ipynb\">GTSRB3</a>\n",
" </span> <a href=\"GTSRB/03-Tracking-and-visualizing.ipynb\">CNN with GTSRB dataset - Monitoring </a><br>\n",
" <span class=\"fid_desc\">Episode 3 : Monitoring and analysing training, managing checkpoints</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/04-Data-augmentation.ipynb\">GTSRB4</a>\n",
" </span> <a href=\"GTSRB/04-Data-augmentation.ipynb\">CNN with GTSRB dataset - Data augmentation </a><br>\n",
" <span class=\"fid_desc\">Episode 4 : Improving the results with data augmentation</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/05-Full-convolutions.ipynb\">GTSRB5</a>\n",
" </span> <a href=\"GTSRB/05-Full-convolutions.ipynb\">CNN with GTSRB dataset - Full convolutions </a><br>\n",
" <span class=\"fid_desc\">Episode 5 : A lot of models, a lot of datasets and a lot of results.</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/06-Notebook-as-a-batch.ipynb\">GTSRB6</a>\n",
" </span> <a href=\"GTSRB/06-Notebook-as-a-batch.ipynb\">Full convolutions as a batch</a><br>\n",
" <span class=\"fid_desc\">Episode 6 : Run Full convolution notebook as a batch</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/07-Show-report.ipynb\">GTSRB7</a>\n",
" </span> <a href=\"GTSRB/07-Show-report.ipynb\">CNN with GTSRB dataset - Show reports</a><br>\n",
" <span class=\"fid_desc\">Episode 7 : Displaying a jobs report</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/batch_oar.sh\">GTSRB10</a>\n",
" </span> <a href=\"GTSRB/batch_oar.sh\">OAR batch submission</a><br>\n",
" <span class=\"fid_desc\">Bash script for OAR batch submission of GTSRB notebook </span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"GTSRB/batch_slurm.sh\">GTSRB11</a>\n",
" </span> <a href=\"GTSRB/batch_slurm.sh\">SLURM batch script</a><br>\n",
" <span class=\"fid_desc\">Bash script for SLURM batch submission of GTSRB notebooks </span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Sentiment analysis with word embedding</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"IMDB/01-Embedding-Keras.ipynb\">IMDB1</a>\n",
" </span> <a href=\"IMDB/01-Embedding-Keras.ipynb\">Text embedding with IMDB</a><br>\n",
" <span class=\"fid_desc\">A very classical example of word embedding for text classification (sentiment analysis)</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"IMDB/02-Prediction.ipynb\">IMDB2</a>\n",
" </span> <a href=\"IMDB/02-Prediction.ipynb\">Text embedding with IMDB - Reloaded</a><br>\n",
" <span class=\"fid_desc\">Example of reusing a previously saved model</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"IMDB/03-LSTM-Keras.ipynb\">IMDB3</a>\n",
" </span> <a href=\"IMDB/03-LSTM-Keras.ipynb\">Text embedding/LSTM model with IMDB</a><br>\n",
" <span class=\"fid_desc\">Still the same problem, but with a network combining embedding and LSTM</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Time series with Recurrent Neural Network (RNN)</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"SYNOP/01-Preparation-of-data.ipynb\">SYNOP1</a>\n",
" </span> <a href=\"SYNOP/01-Preparation-of-data.ipynb\">Time series with RNN - Preparation of data</a><br>\n",
" <span class=\"fid_desc\">Episode 1 : Data analysis and creation of a usable dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"SYNOP/02-First-predictions.ipynb\">SYNOP2</a>\n",
" </span> <a href=\"SYNOP/02-First-predictions.ipynb\">Time series with RNN - Try a prediction</a><br>\n",
" <span class=\"fid_desc\">Episode 2 : Training session and first predictions</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"SYNOP/03-12h-predictions.ipynb\">SYNOP3</a>\n",
" </span> <a href=\"SYNOP/03-12h-predictions.ipynb\">Time series with RNN - 12h predictions</a><br>\n",
" <span class=\"fid_desc\">Episode 3: Attempt to predict in the longer term </span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Unsupervised learning with an autoencoder neural network (AE)</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"AE/01-AE-with-MNIST.ipynb\">AE1</a>\n",
" </span> <a href=\"AE/01-AE-with-MNIST.ipynb\">AutoEncoder (AE) with MNIST</a><br>\n",
" <span class=\"fid_desc\">Episode 1 : Model construction and Training</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"AE/02-AE-with-MNIST-post.ipynb\">AE2</a>\n",
" </span> <a href=\"AE/02-AE-with-MNIST-post.ipynb\">AutoEncoder (AE) with MNIST - Analysis</a><br>\n",
" <span class=\"fid_desc\">Episode 2 : Exploring our denoiser</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Generative network with Variational Autoencoder (VAE)</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/01-VAE-with-MNIST.ipynb\">VAE1</a>\n",
" </span> <a href=\"VAE/01-VAE-with-MNIST.ipynb\">Variational AutoEncoder (VAE) with MNIST</a><br>\n",
" <span class=\"fid_desc\">Building a simple model with the MNIST dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/02-VAE-with-MNIST-post.ipynb\">VAE2</a>\n",
" </span> <a href=\"VAE/02-VAE-with-MNIST-post.ipynb\">Variational AutoEncoder (VAE) with MNIST - Analysis</a><br>\n",
" <span class=\"fid_desc\">Visualization and analysis of latent space</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/05-About-CelebA.ipynb\">VAE3</a>\n",
" </span> <a href=\"VAE/05-About-CelebA.ipynb\">About the CelebA dataset</a><br>\n",
" <span class=\"fid_desc\">Presentation of the CelebA dataset and problems related to its size</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/06-Prepare-CelebA-datasets.ipynb\">VAE6</a>\n",
" </span> <a href=\"VAE/06-Prepare-CelebA-datasets.ipynb\">Preparation of the CelebA dataset</a><br>\n",
" <span class=\"fid_desc\">Preparation of a clustered dataset, batchable</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/07-Check-CelebA.ipynb\">VAE7</a>\n",
" </span> <a href=\"VAE/07-Check-CelebA.ipynb\">Checking the clustered CelebA dataset</a><br>\n",
" <span class=\"fid_desc\">Check the clustered dataset</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/08-VAE-with-CelebA==1090048==.ipynb\">VAE8</a>\n",
" </span> <a href=\"VAE/08-VAE-with-CelebA==1090048==.ipynb\">Variational AutoEncoder (VAE) with CelebA (small)</a><br>\n",
" <span class=\"fid_desc\">Variational AutoEncoder (VAE) with CelebA (small res. 128x128)</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/09-VAE-withCelebA-post.ipynb\">VAE9</a>\n",
" </span> <a href=\"VAE/09-VAE-withCelebA-post.ipynb\">Variational AutoEncoder (VAE) with CelebA - Analysis</a><br>\n",
" <span class=\"fid_desc\">Exploring latent space of our trained models</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"VAE/batch_slurm.sh\">VAE10</a>\n",
" </span> <a href=\"VAE/batch_slurm.sh\">SLURM batch script</a><br>\n",
" <span class=\"fid_desc\">Bash script for SLURM batch submission of VAE notebooks </span>\n",
" </div>\n",
" \n",
"<div class=\"fid_section\">Miscellaneous</div>\n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"Misc/Activation-Functions.ipynb\">ACTF1</a>\n",
" </span> <a href=\"Misc/Activation-Functions.ipynb\">Activation functions</a><br>\n",
" <span class=\"fid_desc\">Some activation functions, with their derivatives.</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"Misc/Numpy.ipynb\">NP1</a>\n",
" </span> <a href=\"Misc/Numpy.ipynb\">A short introduction to Numpy</a><br>\n",
" <span class=\"fid_desc\">Numpy is an essential tool for the Scientific Python.</span>\n",
" </div>\n",
" \n",
"<div class=\"fid_line\">\n",
" <span class=\"fid_id\">\n",
" <a href=\"Misc/Using-Tensorboard.ipynb\">TSB1</a>\n",
" </span> <a href=\"Misc/Using-Tensorboard.ipynb\">Tensorboard with/from Jupyter </a><br>\n",
" <span class=\"fid_desc\">4 ways to use Tensorboard from the Jupyter environment</span>\n",
" </div>\n",
" \n",
"<!-- INDEX_END -->\n",
"\n",
"\n",
"## Installation\n",
"\n",
"A procedure for **configuring** and **starting Jupyter** is available in the **[Wiki](https://gricad-gitlab.univ-grenoble-alpes.fr/talks/fidle/-/wikis/Install-Fidle)**.\n",
"\n",
"## Licence\n",
"\n",
"[<img width=\"100px\" src=\"fidle/img/00-fidle-CC BY-NC-SA.svg\"></img>](https://creativecommons.org/licenses/by-nc-sa/4.0/) \n",
"\\[en\\] Attribution - NonCommercial - ShareAlike 4.0 International (CC BY-NC-SA 4.0) \n",
"\\[Fr\\] Attribution - Pas d’Utilisation Commerciale - Partage dans les Mêmes Conditions 4.0 International \n",
"See [License](https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode). \n",
"See [Disclaimer](https://creativecommons.org/licenses/by-nc-sa/4.0/#). \n",
"\n",
"\n",
"----\n",
"[<img width=\"80px\" src=\"fidle/img/00-Fidle-logo-01.svg\"></img>](#top)\n"
],
"text/plain": [
"<IPython.core.display.Markdown object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from IPython.display import display,Markdown\n",
"display(Markdown(open('README.md', 'r').read()))\n",
"#\n",
"# This README is visible under Jupiter LAb ! :-)"
]
}
],
"metadata": {
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.9"
}
},
"nbformat": 4,
"nbformat_minor": 4
}