diff --git a/MyIA.AI.Notebooks/ML/DataScienceWithAgents/03-DeepLearning/3.1-Retropropagation-From-Scratch.ipynb b/MyIA.AI.Notebooks/ML/DataScienceWithAgents/03-DeepLearning/3.1-Retropropagation-From-Scratch.ipynb deleted file mode 100644 index 8f33d63ee6..0000000000 --- a/MyIA.AI.Notebooks/ML/DataScienceWithAgents/03-DeepLearning/3.1-Retropropagation-From-Scratch.ipynb +++ /dev/null @@ -1,1254 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "id": "e0938739", - "metadata": { - "papermill": { - "duration": 0.005755, - "end_time": "2026-08-23T01:25:57.309313+00:00", - "exception": false, - "start_time": "2026-08-23T01:25:57.303558+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "# 3.1 — Rétropropagation from scratch : la chaîne des gradients à la main\n", - "\n", - "**Navigation** : [<< 2.9-Grokking-Generalisation](../02-ML-Cours/2.9-Grokking-Generalisation.ipynb) | [Index](../README.md)\n", - "\n", - "**Kernel** : Python 3\n", - "\n", - "## Introduction\n", - "\n", - "Le notebook 2.2 a fait descendre un coût le long de sa dérivée : c'est la descente de gradient. Un réseau de neurones fait exactement la même chose — sauf que ses paramètres se comptent ici en dizaines, en millions en pratique, et que la dérivée du coût par rapport à *chacun* d'eux exige de remonter le réseau couche par couche. Cette remontée, c'est la **rétropropagation** : la règle de dérivation en chaîne (*chain rule*) appliquée méthodiquement de la sortie vers l'entrée. C'est elle que PyTorch exécute en silence à chaque appel de `loss.backward()`.\n", - "\n", - "Pourquoi l'écrire à la main, alors qu'autograd existe ? Parce que « le réseau apprend » reste une boîte noire tant qu'on n'a pas écrit, une fois, le gradient à la main — et surtout vérifié qu'il est **juste** par différence finie. Après ce notebook, `backward()` ne sera plus une formule magique : ce sera une ligne de code que vous avez écrite vous-mêmes, et qu'on comparera trait pour trait à PyTorch.\n", - "\n", - "Ce notebook ouvre la sous-série **03-DeepLearning**, en prolongement direct de 02-ML-Cours : même style (tout from scratch d'abord, la bibliothèque ensuite), un jeu de données jouet, une seule nouvelle idée — le gradient qui se propage.\n", - "\n", - "### Objectifs d'apprentissage\n", - "\n", - "À la fin de ce notebook, vous saurez :\n", - "1. Dériver et implémenter la passe avant **et** la passe arrière d'un MLP 2 → 8 → 1 en **NumPy pur** (aucun PyTorch pour apprendre).\n", - "2. **Vérifier un gradient calculé à la main par différence finie** — écart attendu < 1e-7 : la preuve numérique que la dérivation est juste.\n", - "3. Entraîner le réseau sur `make_moons` et visualiser la frontière de décision qui se courbe au fil des itérations.\n", - "4. Montrer que PyTorch (`nn.Linear` + autograd + `SGD`) reproduit **la même trajectoire et les mêmes poids** que notre implémentation.\n", - "5. Diagnostiquer le départ raté de l'initialisation à zéros (symétrie, col, échappée tardive) face à l'initialisation de Xavier.\n", - "\n", - "### Prérequis\n", - "\n", - "- Notebook 2.2 (descente de gradient, dérivée, learning rate).\n", - "- Notebook 2.3 (régression logistique, sigmoïde, entropie croisée).\n", - "- Notebook 2.5b (la BCE est la même perte que la log-vraisemblance de la régression logistique).\n", - "\n", - "> **Référence.** Rumelhart, Hinton & Williams (1986), *Learning representations by back-propagating errors*, Nature 323 — l'article qui a popularisé la rétropropagation.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "id": "b89ab181", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:25:57.318353Z", - "iopub.status.busy": "2026-08-23T01:25:57.316824Z", - "iopub.status.idle": "2026-08-23T01:26:03.064214Z", - "shell.execute_reply": "2026-08-23T01:26:03.064214Z" - }, - "papermill": { - "duration": 5.751912, - "end_time": "2026-08-23T01:26:03.065225+00:00", - "exception": false, - "start_time": "2026-08-23T01:25:57.313313+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Configuration OK : 3.1 - Retropropagation from scratch\n", - "Entrainement : 210 lignes | Test : 90 lignes\n", - "Dimensions de X : (300, 2) (2 variables), classes presentes : [0, 1]\n", - "Classes equilibrees dans le train : [105, 105]\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "torch 2.6.0+cu124 | CUDA disponible : True (inutilise ici, voir section 5)\n" - ] - } - ], - "source": [ - "import warnings\n", - "\n", - "def _warn_no_path(message, category, filename, lineno, file=None, line=None):\n", - " return f\"{category.__name__}: {message}\\n\"\n", - "\n", - "warnings.formatwarning = _warn_no_path\n", - "# Configuration et imports pour le notebook 3.1\n", - "import numpy as np\n", - "import matplotlib.pyplot as plt\n", - "import torch\n", - "\n", - "from sklearn.datasets import make_moons\n", - "from sklearn.model_selection import train_test_split\n", - "from sklearn.linear_model import LogisticRegression\n", - "\n", - "np.random.seed(42)\n", - "torch.manual_seed(42)\n", - "\n", - "# Deux demi-lunes imbriquees : le jouet classique de la classification non lineaire\n", - "X, y = make_moons(n_samples=300, noise=0.2, random_state=42)\n", - "X_train, X_test, y_train, y_test = train_test_split(\n", - " X, y, test_size=0.3, random_state=42, stratify=y\n", - ")\n", - "\n", - "print(\"Configuration OK : 3.1 - Retropropagation from scratch\")\n", - "print(f\"Entrainement : {X_train.shape[0]} lignes | Test : {X_test.shape[0]} lignes\")\n", - "print(f\"Dimensions de X : {X.shape} (2 variables), classes presentes : {np.unique(y).tolist()}\")\n", - "print(f\"Classes equilibrees dans le train : {np.bincount(y_train).tolist()}\")\n", - "print(f\"torch {torch.__version__} | CUDA disponible : {torch.cuda.is_available()} (inutilise ici, voir section 5)\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "7392dd15", - "metadata": { - "papermill": { - "duration": 0.004219, - "end_time": "2026-08-23T01:26:03.073447+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.069228+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 1. Le problème : deux lunes qu'aucune droite ne peut séparer\n", - "\n", - "Le jeu de données `make_moons` dessine deux arcs imbriqués — deux « lunes » face à face. C'est le contre-exemple canonique du modèle linéaire : la régression logistique du notebook 2.3 trace une **droite** comme frontière de décision, et aucune droite ne peut isoler une lune qui enserre l'autre. Le meilleur séparateur linéaire laisse forcément des points des deux classes de chaque côté.\n", - "\n", - "La figure ci-dessous confronte les données (panneau de gauche) à la meilleure frontière linéaire, apprise par `LogisticRegression` (panneau de droite) : la droite coupe les lunes en travers et l'accuracy plafonne. Pour faire mieux, il faut un modèle capable de **courber** sa frontière : un perceptron multicouche (MLP) avec une couche cachée.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "id": "bac977bb", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:03.082985Z", - "iopub.status.busy": "2026-08-23T01:26:03.082985Z", - "iopub.status.idle": "2026-08-23T01:26:03.470680Z", - "shell.execute_reply": "2026-08-23T01:26:03.470680Z" - }, - "papermill": { - "duration": 0.394717, - "end_time": "2026-08-23T01:26:03.471690+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.076973+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Frontiere lineaire : acc train = 0.867 | acc test = 0.833\n", - "Aucune droite ne fait mieux sur ces donnees -> il faut courber la frontiere, donc une couche cachee.\n" - ] - } - ], - "source": [ - "# Figure 1 : les donnees, puis la meilleure frontiere LINEAIRE (regression logistique de 2.3)\n", - "logreg = LogisticRegression().fit(X_train, y_train)\n", - "acc_lin_train = logreg.score(X_train, y_train)\n", - "acc_lin_test = logreg.score(X_test, y_test)\n", - "\n", - "fig, axes = plt.subplots(1, 2, figsize=(11, 4.4))\n", - "\n", - "axes[0].scatter(X_train[y_train == 0, 0], X_train[y_train == 0, 1], color=\"#c44e52\", s=18, label=\"classe 0 (train)\")\n", - "axes[0].scatter(X_train[y_train == 1, 0], X_train[y_train == 1, 1], color=\"#4c72b0\", s=18, label=\"classe 1 (train)\")\n", - "axes[0].scatter(X_test[:, 0], X_test[:, 1], color=\"gray\", s=14, marker=\"x\", label=\"test\")\n", - "axes[0].set_xlabel(\"$x_1$\"); axes[0].set_ylabel(\"$x_2$\")\n", - "axes[0].set_title(\"make_moons : deux lunes imbriquees\")\n", - "axes[0].legend(loc=\"upper right\", fontsize=9)\n", - "\n", - "x_min, x_max = X[:, 0].min() - 0.6, X[:, 0].max() + 0.6\n", - "y_min, y_max = X[:, 1].min() - 0.6, X[:, 1].max() + 0.6\n", - "xx, yy = np.meshgrid(np.linspace(x_min, x_max, 200), np.linspace(y_min, y_max, 200))\n", - "grid = np.c_[xx.ravel(), yy.ravel()]\n", - "proba_lin = logreg.predict_proba(grid)[:, 1].reshape(xx.shape)\n", - "axes[1].contourf(xx, yy, proba_lin, levels=np.linspace(0, 1, 21), cmap=\"RdBu\", alpha=0.55)\n", - "axes[1].contour(xx, yy, proba_lin, levels=[0.5], colors=\"k\", linewidths=1.5)\n", - "axes[1].scatter(X_test[y_test == 0, 0], X_test[y_test == 0, 1], color=\"#c44e52\", s=14)\n", - "axes[1].scatter(X_test[y_test == 1, 0], X_test[y_test == 1, 1], color=\"#4c72b0\", s=14)\n", - "axes[1].set_xlabel(\"$x_1$\"); axes[1].set_ylabel(\"$x_2$\")\n", - "axes[1].set_title(f\"Meilleure droite (logistique) : acc test = {acc_lin_test:.3f}\")\n", - "\n", - "fig.tight_layout()\n", - "plt.show()\n", - "\n", - "print(f\"Frontiere lineaire : acc train = {acc_lin_train:.3f} | acc test = {acc_lin_test:.3f}\")\n", - "print(\"Aucune droite ne fait mieux sur ces donnees -> il faut courber la frontiere, donc une couche cachee.\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "bd46767a", - "metadata": { - "papermill": { - "duration": 0.005032, - "end_time": "2026-08-23T01:26:03.482858+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.477826+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "### L'architecture 2 → 8 → 1 et la passe avant\n", - "\n", - "Un MLP minimal à une couche cachée, en notations matricielles (N exemples en ligne) :\n", - "\n", - "- $Z_1 = X W_1 + b_1$ de taille $(N, 8)$, puis $A_1 = \\sigma(Z_1)$ : **8 neurones cachés**, chacun une sigmoïde d'une combinaison linéaire des 2 entrées ;\n", - "- $Z_2 = A_1 W_2 + b_2$ de taille $(N, 1)$, puis $\\hat{y} = \\sigma(Z_2)$ : **une probabilité de sortie**.\n", - "\n", - "Soit quatre tenseurs de paramètres — $W_1$ $(2 \\times 8)$, $b_1$ $(8)$, $W_2$ $(8 \\times 1)$, $b_2$ $(1)$ — **33 paramètres** au total. Chaque neurone caché apprend son propre « détecteur » (une petite frontière orientée dans le plan), et la sortie agrège les 8 détecteurs en une probabilité. La **passe avant** n'est que ces deux produits matriciels : le réseau non entraîné sait déjà l'exécuter — il prédit simplement n'importe quoi.\n", - "\n", - "L'initialisation : chaque poids est tiré selon $\\mathcal{N}(0, 1/\\mathrm{fan\\_in})$ (Xavier, calibrée pour la sigmoïde — la section 6 montre pourquoi le choix importe), les biais à zéro.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "id": "0b080c91", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:03.494771Z", - "iopub.status.busy": "2026-08-23T01:26:03.493762Z", - "iopub.status.idle": "2026-08-23T01:26:03.506344Z", - "shell.execute_reply": "2026-08-23T01:26:03.505830Z" - }, - "papermill": { - "duration": 0.020132, - "end_time": "2026-08-23T01:26:03.507349+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.487217+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Parametres : W1 (2, 8), b1 (8,), W2 (8, 1), b2 (1,) -> 33 au total\n", - "Reseau non entraine : loss BCE (test) = 0.6781\n", - "Accuracy test d'un reseau non entraine : 0.589 -- autant tirer a pile ou face\n" - ] - } - ], - "source": [ - "# Le MLP 2 -> 8 -> 1 en NumPy pur : initialisation Xavier + passe avant\n", - "H = 8 # largeur de la couche cachee\n", - "\n", - "def sigmoid(z):\n", - " return 1.0 / (1.0 + np.exp(-z))\n", - "\n", - "def init_params(seed=42, mode=\"xavier\"):\n", - " \"\"\"Initialise (W1, b1, W2, b2).\n", - "\n", - " mode=\"xavier\" : poids ~ N(0, 1/fan_in), variance calibree pour la sigmoide\n", - " (Glorot & Bengio, 2010). mode=\"zeros\" : tout a zero (section 6).\n", - " \"\"\"\n", - " rng = np.random.default_rng(seed)\n", - " d = X.shape[1]\n", - " if mode == \"xavier\":\n", - " W1 = rng.normal(0.0, np.sqrt(1.0 / d), size=(d, H))\n", - " W2 = rng.normal(0.0, np.sqrt(1.0 / H), size=(H, 1))\n", - " else: # zeros\n", - " W1 = np.zeros((d, H))\n", - " W2 = np.zeros((H, 1))\n", - " b1 = np.zeros(H)\n", - " b2 = np.zeros(1)\n", - " return W1, b1, W2, b2\n", - "\n", - "def forward(params, Xd):\n", - " \"\"\"Passe avant : retourne (y_hat, cache). Le cache sert a la passe arriere.\"\"\"\n", - " W1, b1, W2, b2 = params\n", - " z1 = Xd @ W1 + b1 # (N, 8)\n", - " a1 = sigmoid(z1) # (N, 8)\n", - " z2 = a1 @ W2 + b2 # (N, 1)\n", - " y_hat = sigmoid(z2) # (N, 1)\n", - " return y_hat, (z1, a1, z2)\n", - "\n", - "def bce_loss(y_hat, y):\n", - " \"\"\"Entropie croisee binaire moyenne (la meme perte que la regression logistique 2.3).\"\"\"\n", - " y_hat = np.clip(y_hat, 1e-12, 1.0 - 1e-12)\n", - " y = y.reshape(-1, 1)\n", - " return float(np.mean(-(y * np.log(y_hat) + (1.0 - y) * np.log(1.0 - y_hat))))\n", - "\n", - "params = init_params(seed=42, mode=\"xavier\")\n", - "y_hat_test, _ = forward(params, X_test)\n", - "acc_init = ((y_hat_test > 0.5).astype(int).ravel() == y_test).mean()\n", - "\n", - "print(f\"Parametres : W1 {params[0].shape}, b1 {params[1].shape}, W2 {params[2].shape}, b2 {params[3].shape}\"\n", - " f\" -> {sum(p.size for p in params)} au total\")\n", - "print(f\"Reseau non entraine : loss BCE (test) = {bce_loss(y_hat_test, y_test):.4f}\")\n", - "print(f\"Accuracy test d'un reseau non entraine : {acc_init:.3f} -- autant tirer a pile ou face\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "0591b9ae", - "metadata": { - "papermill": { - "duration": 0.00552, - "end_time": "2026-08-23T01:26:03.517964+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.512444+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 2. La passe arrière à la main : la chain rule couche par couche\n", - "\n", - "Apprendre = ajuster les 33 paramètres dans le sens opposé du gradient de la loss. La loss est la BCE (entropie croisée binaire moyenne — celle de la régression logistique du notebook 2.3). Il faut donc $\\partial L / \\partial W_1$, $\\partial L / \\partial b_1$, $\\partial L / \\partial W_2$, $\\partial L / \\partial b_2$. La chain rule les construit en remontant, en quatre étapes.\n", - "\n", - "**Étape 1 — la sortie : BCE et sigmoïde se compensent.** Côté loss, $\\frac{\\partial L}{\\partial \\hat{y}} = \\frac{\\hat{y} - y}{\\hat{y}(1 - \\hat{y})}$ ; côté activation, $\\sigma'(z_2) = \\hat{y}(1 - \\hat{y})$. Le produit des deux se simplifie en\n", - "\n", - "$$\\delta_2 \\equiv \\frac{\\partial L}{\\partial z_2} = \\hat{y} - y$$\n", - "\n", - "— littéralement *l'erreur de prédiction*. Cette simplification est la raison pour laquelle on couple toujours BCE et sigmoïde : sans elle, le facteur $\\sigma'$ s'écraserait vers 0 quand $\\hat{y}$ approche 0 ou 1, et le gradient s'éteindrait.\n", - "\n", - "**Étape 2 — gradients de la dernière couche** (règle de dimensions : le gradient d'un paramètre a la forme du paramètre) :\n", - "\n", - "$$\\frac{\\partial L}{\\partial W_2} = A_1^\\top \\delta_2, \\qquad \\frac{\\partial L}{\\partial b_2} = \\sum_n \\delta_2^{(n)}$$\n", - "\n", - "avec le facteur $1/N$ de la moyenne rangé dans $\\delta_2$ dans le code ci-dessous.\n", - "\n", - "**Étape 3 — remonter à travers $W_2$ vers la couche cachée :**\n", - "\n", - "$$\\frac{\\partial L}{\\partial A_1} = \\delta_2 W_2^\\top, \\qquad \\delta_1 \\equiv \\frac{\\partial L}{\\partial Z_1} = \\frac{\\partial L}{\\partial A_1} \\odot A_1 (1 - A_1)$$\n", - "\n", - "où $\\odot$ est le produit terme à terme et $A_1 \\odot (1 - A_1)$ est $\\sigma'(Z_1)$ réécrit sans $Z_1$ — c'est à ça que sert le *cache* de la passe avant.\n", - "\n", - "**Étape 4 — gradients de la première couche :**\n", - "\n", - "$$\\frac{\\partial L}{\\partial W_1} = X^\\top \\delta_1, \\qquad \\frac{\\partial L}{\\partial b_1} = \\sum_n \\delta_1^{(n)}$$\n", - "\n", - "Le motif « $\\delta$ → gradients → nouveau $\\delta$ » est **le même à chaque couche** : empiler des couches ne fait qu'empiler des étapes 2–3. C'est toute la rétropropagation.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "id": "ea8496c6", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:03.529646Z", - "iopub.status.busy": "2026-08-23T01:26:03.529646Z", - "iopub.status.idle": "2026-08-23T01:26:03.537579Z", - "shell.execute_reply": "2026-08-23T01:26:03.536568Z" - }, - "papermill": { - "duration": 0.014607, - "end_time": "2026-08-23T01:26:03.537579+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.522972+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Formes des gradients (doivent egaler celles des parametres) :\n", - " dW1 (2, 8) (parametre (2, 8) ) | |grad| moyen = 9.06e-03\n", - " db1 (8,) (parametre (8,) ) | |grad| moyen = 2.43e-03\n", - " dW2 (8, 1) (parametre (8, 1) ) | |grad| moyen = 3.85e-02\n", - " db2 (1,) (parametre (1,) ) | |grad| moyen = 4.72e-02\n" - ] - } - ], - "source": [ - "# Passe arriere en NumPy pur : la chain rule, couche par couche\n", - "def backward(params, Xd, y, cache):\n", - " \"\"\"Retourne (dW1, db1, dW2, db2) : gradient de la BCE moyenne pour chaque parametre.\"\"\"\n", - " W1, b1, W2, b2 = params\n", - " z1, a1, z2 = cache\n", - " N = Xd.shape[0]\n", - " y_col = y.reshape(-1, 1)\n", - "\n", - " dz2 = (sigmoid(z2) - y_col) / N # (N,1) etape 1 : BCE + sigmoide -> y_hat - y (facteur 1/N range ici)\n", - " dW2 = a1.T @ dz2 # (8,1) etape 2\n", - " db2 = dz2.sum(axis=0) # (1,)\n", - " da1 = dz2 @ W2.T # (N,8) etape 3 : remonter a travers W2\n", - " dz1 = da1 * a1 * (1.0 - a1) # (N,8) etape 3 : x sigmoide'(z1) = a1 (1 - a1)\n", - " dW1 = Xd.T @ dz1 # (2,8) etape 4\n", - " db1 = dz1.sum(axis=0) # (8,)\n", - " return dW1, db1, dW2, db2\n", - "\n", - "y_hat_train, cache = forward(params, X_train)\n", - "grads = backward(params, X_train, y_train, cache)\n", - "\n", - "print(\"Formes des gradients (doivent egaler celles des parametres) :\")\n", - "for name, p, g in zip([\"W1\", \"b1\", \"W2\", \"b2\"], params, grads):\n", - " print(f\" d{name} {str(g.shape):8s} (parametre {str(p.shape):8s}) | |grad| moyen = {np.abs(g).mean():.2e}\")\n" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "id": "11a1f90a", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:03.553888Z", - "iopub.status.busy": "2026-08-23T01:26:03.553888Z", - "iopub.status.idle": "2026-08-23T01:26:03.564200Z", - "shell.execute_reply": "2026-08-23T01:26:03.563173Z" - }, - "papermill": { - "duration": 0.017867, - "end_time": "2026-08-23T01:26:03.564200+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.546333+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - " indice grad analytique grad numerique ecart\n", - " 3 7.921e-04 7.921e-04 3.1e-12\n", - " 12 -2.714e-03 -2.714e-03 1.0e-11\n", - " 19 -2.178e-04 -2.178e-04 1.3e-11\n", - " 27 1.649e-02 1.649e-02 3.0e-12\n", - " 32 4.718e-02 4.718e-02 8.0e-13\n", - "\n", - "Ecart max analytique vs numerique : 1.33e-11 (seuil 1e-7) -> VERIFIE : la derivation manuelle est juste\n" - ] - } - ], - "source": [ - "# VERIFICATION du gradient par difference finie : la preuve numerique que la derivation est juste\n", - "def flatten(pl):\n", - " return np.concatenate([p.ravel() for p in pl])\n", - "\n", - "def unflatten(theta):\n", - " \"\"\"Reconstruit (W1, b1, W2, b2) depuis le vecteur plat theta.\"\"\"\n", - " formes = [(X.shape[1], H), (H,), (H, 1), (1,)]\n", - " out, i = [], 0\n", - " for f in formes:\n", - " n = int(np.prod(f))\n", - " out.append(theta[i:i + n].reshape(f))\n", - " i += n\n", - " return tuple(out)\n", - "\n", - "def loss_at(theta, Xd, y):\n", - " y_hat, _ = forward(unflatten(theta), Xd)\n", - " return bce_loss(y_hat, y)\n", - "\n", - "theta = flatten(params)\n", - "grad_flat = flatten(grads)\n", - "eps = 1e-5\n", - "idx_choisis = [3, 12, 19, 27, 32] # pioches dans W1, W1, b1, W2 et b2\n", - "\n", - "print(f\"{'indice':>7} {'grad analytique':>17} {'grad numerique':>17} {'ecart':>10}\")\n", - "ecarts = []\n", - "for i in idx_choisis:\n", - " tp, tm = theta.copy(), theta.copy()\n", - " tp[i] += eps\n", - " tm[i] -= eps\n", - " g_num = (loss_at(tp, X_train, y_train) - loss_at(tm, X_train, y_train)) / (2 * eps)\n", - " ecart = abs(grad_flat[i] - g_num)\n", - " ecarts.append(ecart)\n", - " print(f\"{i:>7} {grad_flat[i]:>17.3e} {g_num:>17.3e} {ecart:>10.1e}\")\n", - "\n", - "max_ecart = max(ecarts)\n", - "verdict = \"VERIFIE : la derivation manuelle est juste\" if max_ecart < 1e-7 else \"PROBLEME : revoir la derivation\"\n", - "print(f\"\\nEcart max analytique vs numerique : {max_ecart:.2e} (seuil 1e-7) -> {verdict}\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "b98617ca", - "metadata": { - "papermill": { - "duration": 0.004989, - "end_time": "2026-08-23T01:26:03.574186+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.569197+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 3. La boucle d'entraînement : suivre le gradient, 20 000 fois\n", - "\n", - "Il ne reste qu'à assembler : à chaque itération, passe avant (on garde le cache), passe arrière (les 4 gradients), puis un pas de descente $\\theta \\leftarrow \\theta - \\eta \\nabla_\\theta L$ pour chacun des 33 paramètres. Learning rate $\\eta = 0.5$, 20 000 itérations (quelques secondes de CPU — le full-batch NumPy sur 210 exemples est très rapide), **full-batch** : tout le train sert à chaque pas, comme au notebook 2.2 — pas de mini-batchs, le mécanisme à nu.\n", - "\n", - "Deux lectures à préparer. La loss BCE vaut $\\ln 2 \\approx 0{,}693$ pour des prédictions au hasard ; nettement en dessous de 0.10, le réseau est confiant et rarement faux. Et la frontière de décision raconte la même histoire en image : nous prenons des instantanés aux itérations 0, 1 000, 5 000 et 20 000.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "id": "cc5d8134", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:03.583930Z", - "iopub.status.busy": "2026-08-23T01:26:03.583930Z", - "iopub.status.idle": "2026-08-23T01:26:05.212909Z", - "shell.execute_reply": "2026-08-23T01:26:05.211905Z" - }, - "papermill": { - "duration": 1.634221, - "end_time": "2026-08-23T01:26:05.212909+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:03.578688+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Loss BCE (train) : 0.6699 -> 0.0529 en 20000 iterations\n", - "Accuracy test du reseau entraine : 0.944\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Boucle d'entrainement full-batch : descente de gradient toute simple\n", - "def train_mlp(params, Xd, y, lr=0.5, iters=5000, snapshots_at=()):\n", - " \"\"\"Descente full-batch. Retourne (params_finaux, loss_history, {iteration: params}).\"\"\"\n", - " loss_history = np.empty(iters)\n", - " snaps = {0: tuple(p.copy() for p in params)}\n", - " for it in range(iters):\n", - " y_hat, cache = forward(params, Xd)\n", - " loss_history[it] = bce_loss(y_hat, y)\n", - " dW1, db1, dW2, db2 = backward(params, Xd, y, cache)\n", - " params = (params[0] - lr * dW1, params[1] - lr * db1,\n", - " params[2] - lr * dW2, params[3] - lr * db2)\n", - " if (it + 1) in snapshots_at:\n", - " snaps[it + 1] = tuple(p.copy() for p in params)\n", - " return params, loss_history, snaps\n", - "\n", - "params_train, loss_hist, snaps = train_mlp(\n", - " init_params(seed=42, mode=\"xavier\"), X_train, y_train,\n", - " lr=0.5, iters=20000, snapshots_at=(1000, 5000, 20000),\n", - ")\n", - "\n", - "y_hat_test, _ = forward(params_train, X_test)\n", - "acc_test = ((y_hat_test > 0.5).astype(int).ravel() == y_test).mean()\n", - "\n", - "print(f\"Loss BCE (train) : {loss_hist[0]:.4f} -> {loss_hist[-1]:.4f} en 20000 iterations\")\n", - "print(f\"Accuracy test du reseau entraine : {acc_test:.3f}\")\n", - "\n", - "plt.figure(figsize=(7.2, 4.2))\n", - "plt.plot(loss_hist, color=\"#4c72b0\", linewidth=2)\n", - "plt.axhline(np.log(2), color=\"gray\", linestyle=\"--\", linewidth=1, label=\"hasard complet (ln 2 = 0.693)\")\n", - "plt.xlabel(\"Iteration\"); plt.ylabel(\"Loss BCE (train)\")\n", - "plt.title(\"Descente de la loss pendant l'entrainement\")\n", - "plt.legend(); plt.grid(alpha=0.3); plt.tight_layout(); plt.show()\n" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "id": "10b9f1ee", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:05.226912Z", - "iopub.status.busy": "2026-08-23T01:26:05.226912Z", - "iopub.status.idle": "2026-08-23T01:26:05.766912Z", - "shell.execute_reply": "2026-08-23T01:26:05.765905Z" - }, - "papermill": { - "duration": 0.54751, - "end_time": "2026-08-23T01:26:05.767419+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.219909+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# La frontiere de decision a quatre stades de l'apprentissage\n", - "def make_grid(Xd, n=200, pad=0.6):\n", - " xx_, yy_ = np.meshgrid(\n", - " np.linspace(Xd[:, 0].min() - pad, Xd[:, 0].max() + pad, n),\n", - " np.linspace(Xd[:, 1].min() - pad, Xd[:, 1].max() + pad, n),\n", - " )\n", - " return xx_, yy_, np.c_[xx_.ravel(), yy_.ravel()]\n", - "\n", - "def plot_boundary(ax, params, Xd, yd, title, fwd=None):\n", - " \"\"\"Carte de probabilite du reseau + nuage des exemples (frontiere = niveau 0.5).\"\"\"\n", - " if fwd is None:\n", - " fwd = forward\n", - " xx_, yy_, grid_ = make_grid(Xd)\n", - " probs = fwd(params, grid_)[0].reshape(xx_.shape)\n", - " ax.contourf(xx_, yy_, probs, levels=np.linspace(0, 1, 21), cmap=\"RdBu\", alpha=0.55)\n", - " if probs.min() < 0.5 < probs.max():\n", - " ax.contour(xx_, yy_, probs, levels=[0.5], colors=\"k\", linewidths=1.2)\n", - " ax.scatter(Xd[yd == 0, 0], Xd[yd == 0, 1], color=\"#c44e52\", s=14)\n", - " ax.scatter(Xd[yd == 1, 0], Xd[yd == 1, 1], color=\"#4c72b0\", s=14)\n", - " ax.set_title(title, fontsize=11)\n", - " ax.set_xlabel(\"$x_1$\"); ax.set_ylabel(\"$x_2$\")\n", - "\n", - "fig, axes = plt.subplots(2, 2, figsize=(11, 8.2))\n", - "for ax, (it, p) in zip(axes.ravel(), snaps.items()):\n", - " l = bce_loss(forward(p, X_train)[0], y_train)\n", - " plot_boundary(ax, p, X_train, y_train, f\"iteration {it} -- loss train = {l:.3f}\")\n", - "fig.suptitle(\"Frontiere de decision du MLP 2 -> 8 -> 1 pendant l'apprentissage\")\n", - "fig.tight_layout()\n", - "plt.show()\n" - ] - }, - { - "cell_type": "markdown", - "id": "62dd61fe", - "metadata": { - "papermill": { - "duration": 0.008003, - "end_time": "2026-08-23T01:26:05.783943+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.775940+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "### Lecture du résultat\n", - "\n", - "La loss plonge sous le niveau du hasard ($\\ln 2$) en quelques centaines d'itérations, puis continue de s'enfoncer : 0.670 → 0.053, et l'accuracy test passe de **0.833** (la meilleure droite, section 1) à **0.944**. La séquence des frontières dit la même histoire : à l'itération 0, le réseau non entraîné découpe le plan de façon quasi arbitraire ; dès 1 000, la frontière est déjà presque une droite (l'essentiel : séparer globalement haut/bas) ; vers 5 000 puis 20 000, elle **se courbe** pour contourner chaque lune — précisément ce qu'aucune droite ne pouvait faire. Les erreurs restantes (5 points sur 90) se logent dans le chevauchement des deux lunes, là où le bruit du générateur a fait déborder des points dans la classe opposée.\n" - ] - }, - { - "cell_type": "markdown", - "id": "76fab6b2", - "metadata": { - "papermill": { - "duration": 0.008516, - "end_time": "2026-08-23T01:26:05.801998+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.793482+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 4. L'effet du learning rate\n", - "\n", - "La boucle ci-dessus fonctionne avec $\\eta = 0.5$. Que se passe-t-il quand on le change ? Le learning rate règle la taille du pas : trop petit, chaque itération corrige à peine l'erreur (**stagnation**) ; trop grand, le pas franchit le minimum et rebondit de l'autre côté de la vallée, parfois jusqu'à l'explosion (**divergence**) ; bien réglé, descente rapide et stable. C'est la même leçon qu'au notebook 2.2 — mais désormais vous possédez le code pour la mesurer vous-mêmes.\n", - "\n", - "### Exercice 1 : convergence, stagnation ou divergence ?\n", - "\n", - "Entraînez le même réseau (même `seed=42`, même architecture, 2 000 itérations suffisent) avec $\\eta \\in \\{0.01,\\ 0.5,\\ 5.0\\}$, tracez les trois courbes de loss sur la même figure, et **prédisez avant de lancer** lequel diverge.\n", - "\n", - "Indices :\n", - "- `# Étape 1` : `train_mlp(init_params(seed=42), X_train, y_train, lr=..., iters=2000)` — trois fois, un `lr` par courbe.\n", - "- `# Étape 2` : tracer les trois `loss_history` sur la même figure ; `plt.yscale(\"log\")` rend la comparaison lisible.\n", - "- `# Étape 3` : repérer la divergence : une loss finale supérieure à $\\ln 2$ après être montée, ou des valeurs devenues absurdes.\n", - "- Lecture attendue : 0.01 stagne (il faudrait ~50× plus d'itérations pour le même effet) ; 0.5 descend bas et vite ; 5.0 est trop grand — le pas saute par-dessus le minimum.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "id": "396e461d", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:05.820708Z", - "iopub.status.busy": "2026-08-23T01:26:05.819694Z", - "iopub.status.idle": "2026-08-23T01:26:05.825226Z", - "shell.execute_reply": "2026-08-23T01:26:05.824716Z" - }, - "papermill": { - "duration": 0.015591, - "end_time": "2026-08-23T01:26:05.826242+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.810651+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Exercice a completer : 0.01 stagne, 0.5 converge, 5.0 diverge ? Verifiez sur la figure.\n" - ] - } - ], - "source": [ - "# Exercice 1 : effet du learning rate (eta) sur la descente\n", - "losses_eta = {} # TODO etudiant : boucle sur eta dans [0.01, 0.5, 5.0]\n", - "# TODO etudiant : losses_eta[eta] = train_mlp(init_params(seed=42), X_train, y_train, lr=eta, iters=2000)[1]\n", - "# TODO etudiant : tracer les trois courbes sur la meme figure (plt.yscale(\"log\") utile)\n", - "print(\"Exercice a completer : 0.01 stagne, 0.5 converge, 5.0 diverge ? Verifiez sur la figure.\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "d68eb6d4", - "metadata": { - "papermill": { - "duration": 0.007, - "end_time": "2026-08-23T01:26:05.841743+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.834743+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 5. La même chose avec PyTorch : ce qu'autograd fait en silence\n", - "\n", - "Tout ce que nous venons d'écrire — passe avant, cache, chain rule, gradients — PyTorch l'exécute dans `loss.backward()`. Pour le prouver, on construit le même réseau avec `torch.nn.Linear`, la même perte (`BCELoss` sur une sortie sigmoïde), le même optimiseur SGD $\\eta = 0.5$, le même nombre d'itérations, et surtout **exactement la même initialisation** : on copie les poids Xavier NumPy dans les tenseurs PyTorch, et on travaille en `float64` pour éliminer les écarts d'arrondi.\n", - "\n", - "Prédiction : loss identique au premier pas, trajectoires quasi identiques, mêmes poids à convergence. Si c'est bien le cas, notre backward à la main calculait exactement ce qu'autograd calcule — et c'est ce qui rend PyTorch fiable : pas de magie, seulement la chain rule industrialisée. (Le GPU serait inutile ici — 33 paramètres, quelques secondes de CPU ; on privilégie l'**exactitude** avant la vitesse. Le GPU redevient utile en 3.2.)\n" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "id": "21d5f9e7", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:05.858932Z", - "iopub.status.busy": "2026-08-23T01:26:05.858932Z", - "iopub.status.idle": "2026-08-23T01:26:15.951241Z", - "shell.execute_reply": "2026-08-23T01:26:15.950186Z" - }, - "papermill": { - "duration": 10.10251, - "end_time": "2026-08-23T01:26:15.951771+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:05.849261+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Premier pas : NumPy 0.669933 | PyTorch 0.669933\n", - "Dernier pas : NumPy 0.052877 | PyTorch 0.052877\n", - "Ecart max sur les 20000 losses : 2.22e-16\n", - "Poids a convergence identiques (atol=1e-6) : True\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Le MEME reseau en PyTorch : nn.Linear + BCELoss + SGD, initialisation copiee depuis NumPy\n", - "torch.manual_seed(42)\n", - "\n", - "model = torch.nn.Sequential(\n", - " torch.nn.Linear(2, H), torch.nn.Sigmoid(),\n", - " torch.nn.Linear(H, 1), torch.nn.Sigmoid(),\n", - ").double()\n", - "\n", - "W1_0, b1_0, W2_0, b2_0 = init_params(seed=42, mode=\"xavier\") # le meme tirage Xavier que la section 1\n", - "with torch.no_grad():\n", - " model[0].weight.copy_(torch.from_numpy(W1_0.T))\n", - " model[0].bias.copy_(torch.from_numpy(b1_0))\n", - " model[2].weight.copy_(torch.from_numpy(W2_0.T))\n", - " model[2].bias.copy_(torch.from_numpy(b2_0))\n", - "\n", - "Xt = torch.from_numpy(X_train)\n", - "yt = torch.from_numpy(y_train.reshape(-1, 1)).double()\n", - "bce = torch.nn.BCELoss()\n", - "optimizer = torch.optim.SGD(model.parameters(), lr=0.5)\n", - "\n", - "loss_torch = np.empty(20000)\n", - "for it in range(20000):\n", - " optimizer.zero_grad()\n", - " loss = bce(model(Xt), yt)\n", - " loss.backward()\n", - " optimizer.step()\n", - " loss_torch[it] = loss.item()\n", - "\n", - "# Contre-partie NumPy : la run de la section 3 -- init_params(seed=42) est deterministe,\n", - "# donc les poids copies ci-dessus sont EXACTEMENT ses poids de depart (meme lr, meme budget)\n", - "params_np, loss_np = params_train, loss_hist\n", - "\n", - "W1_np, b1_np, W2_np, b2_np = params_np\n", - "meme_poids = (\n", - " np.allclose(W1_np, model[0].weight.detach().numpy().T, atol=1e-6)\n", - " and np.allclose(b1_np, model[0].bias.detach().numpy(), atol=1e-6)\n", - " and np.allclose(W2_np, model[2].weight.detach().numpy().T, atol=1e-6)\n", - " and np.allclose(b2_np, model[2].bias.detach().numpy(), atol=1e-6)\n", - ")\n", - "\n", - "print(f\"Premier pas : NumPy {loss_np[0]:.6f} | PyTorch {loss_torch[0]:.6f}\")\n", - "print(f\"Dernier pas : NumPy {loss_np[-1]:.6f} | PyTorch {loss_torch[-1]:.6f}\")\n", - "print(f\"Ecart max sur les 20000 losses : {np.max(np.abs(loss_np - loss_torch)):.2e}\")\n", - "print(f\"Poids a convergence identiques (atol=1e-6) : {meme_poids}\")\n", - "\n", - "plt.figure(figsize=(7.2, 4.2))\n", - "plt.plot(loss_np, color=\"#4c72b0\", linewidth=2, label=\"NumPy (backward a la main)\")\n", - "plt.plot(loss_torch, color=\"#c44e52\", linewidth=2, linestyle=\"--\", label=\"PyTorch (autograd + SGD)\")\n", - "plt.xlabel(\"Iteration\"); plt.ylabel(\"Loss BCE (train)\")\n", - "plt.title(\"Deux implementations, une seule trajectoire\")\n", - "plt.legend(); plt.grid(alpha=0.3); plt.tight_layout(); plt.show()\n" - ] - }, - { - "cell_type": "markdown", - "id": "577d939d", - "metadata": { - "papermill": { - "duration": 0.009169, - "end_time": "2026-08-23T01:26:15.971060+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:15.961891+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 6. L'effet de l'initialisation : Xavier contre zéros\n", - "\n", - "Reprenez le réseau ci-dessus et mettez tous les poids **et** biais à zéro. Même architecture, même loss, même $\\eta$ — et pendant les premières ~1 700 itérations il ne se passe **strictement rien** : la loss reste collée à $\\ln 2 = 0.693$. La raison : la **symétrie**. Si tous les poids de $W_1$ sont nuls, les 8 neurones cachés reçoivent la même entrée (zéro) et produisent tous $A_1 = 0.5$ ; le gradient de $W_1$ est alors identique colonne après colonne, et la mise à jour maintient les colonnes égales. Et comme les classes sont équilibrées (105 / 105 dans le train), $\\partial L / \\partial b_2 = \\mathrm{moyenne}(0.5 - y) = 0$ **en arithmétique exacte**. Chaque neurone est le sosie de son voisin : en arithmétique exacte, le réseau resterait gelé pour toujours.\n", - "\n", - "Mais ce point symétrique n'est pas un minimum : c'est un **col** (point selle). En float64, la somme $\\mathrm{moyenne}(0.5 - y)$ laisse un résidu d'arrondi de l'ordre de $10^{-18}$ — une poussière, mais une poussière **non nulle**, qui met $b_2$ en mouvement. Itération après itération, cette graine croît d'environ 2 à 3 % par pas (le col est instable dans la direction qui brise la symétrie) : au bout de ~1 700 itérations elle devient visible et le réseau « s'échappe ». L'échappée ne le sauve pas : parti de nulle part, il atterrit dans un bassin plus mauvais — mesurez-le dans la cellule suivante, loss finale ~0.27 contre ~0.05, accuracy ~86 % contre ~94 %.\n", - "\n", - "L'initialisation de Xavier évite tout ce théâtre : chaque poids est tiré indépendamment autour de 0, avec une variance $1/\\mathrm{fan\\_in}$ calibrée pour que les activations ne s'éteignent ni ne saturent à travers les couches (Glorot & Bengio, 2010). La symétrie est cassée dès la première itération, et tout le budget d'entraînement sert à apprendre. Même réseau, même loss, même $\\eta$ — seul le point de départ change, et avec lui tout le reste.\n" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "id": "cb502867", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:15.989866Z", - "iopub.status.busy": "2026-08-23T01:26:15.989866Z", - "iopub.status.idle": "2026-08-23T01:26:17.958031Z", - "shell.execute_reply": "2026-08-23T01:26:17.957200Z" - }, - "papermill": { - "duration": 1.97946, - "end_time": "2026-08-23T01:26:17.959041+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:15.979581+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Init zeros : figee a ln 2 = 0.6931 pendant 1687 iterations, puis echappee tardive\n", - " loss finale 0.2654, accuracy test 0.856\n", - "Init Xavier : loss finale 0.0529, accuracy test 0.944\n", - "Norme de W1 apres 20000 iterations : zeros 3.52 | Xavier 17.61\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Xavier vs zeros : meme architecture, meme loss, meme eta -- seul le depart change\n", - "params_z, loss_z, _ = train_mlp(init_params(seed=42, mode=\"zeros\"), X_train, y_train, lr=0.5, iters=20000)\n", - "params_x, loss_x = params_train, loss_hist # cote Xavier : la run de la section 3 (config identique)\n", - "\n", - "acc_z = ((forward(params_z, X_test)[0] > 0.5).astype(int).ravel() == y_test).mean()\n", - "acc_x = ((forward(params_x, X_test)[0] > 0.5).astype(int).ravel() == y_test).mean()\n", - "it_echappee = int(np.argmax(loss_z < np.log(2) - 1e-6)) # premiere iteration ou la loss quitte ln 2\n", - "\n", - "print(f\"Init zeros : figee a ln 2 = {np.log(2):.4f} pendant {it_echappee} iterations, puis echappee tardive\")\n", - "print(f\" loss finale {loss_z[-1]:.4f}, accuracy test {acc_z:.3f}\")\n", - "print(f\"Init Xavier : loss finale {loss_x[-1]:.4f}, accuracy test {acc_x:.3f}\")\n", - "print(f\"Norme de W1 apres 20000 iterations : zeros {np.linalg.norm(params_z[0]):.2f} | Xavier {np.linalg.norm(params_x[0]):.2f}\")\n", - "\n", - "fig, axes = plt.subplots(1, 3, figsize=(13.5, 4.0))\n", - "axes[0].plot(loss_z, color=\"#c44e52\", linewidth=2, label=\"init zeros\")\n", - "axes[0].plot(loss_x, color=\"#4c72b0\", linewidth=2, label=\"init Xavier\")\n", - "axes[0].axhline(np.log(2), color=\"gray\", linestyle=\"--\", linewidth=1)\n", - "axes[0].set_xlabel(\"Iteration\"); axes[0].set_ylabel(\"Loss BCE (train)\")\n", - "axes[0].set_title(\"Trajectoires de loss\"); axes[0].legend(); axes[0].grid(alpha=0.3)\n", - "plot_boundary(axes[1], params_z, X_train, y_train, f\"zeros -- echappee tardive, bassin plus mauvais (acc {acc_z:.2f})\")\n", - "plot_boundary(axes[2], params_x, X_train, y_train, f\"Xavier -- symetrie cassee d'entree (acc test {acc_x:.2f})\")\n", - "fig.tight_layout(); plt.show()\n" - ] - }, - { - "cell_type": "markdown", - "id": "2b186bf8", - "metadata": { - "papermill": { - "duration": 0.009551, - "end_time": "2026-08-23T01:26:17.978795+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:17.969244+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 7. La profondeur : une couche cachée ou deux ?\n", - "\n", - "Notre backward était écrit pour une couche. La beauté de la chain rule : le motif $\\delta$ → gradients → $\\delta$ se répète à chaque couche, si bien que le code se généralise en une boucle — un réseau devient une simple liste de couches $(W, b)$. Le mini-cas ci-dessous compare $2 \\to 8 \\to 1$ et $2 \\to 8 \\to 8 \\to 1$ sur `make_circles`, deux cercles concentriques.\n", - "\n", - "### Exercice 2 : prédire, puis faire varier\n", - "\n", - "Avant de lancer la cellule : quelle architecture atteindra la meilleure accuracy test ? Lancez, comparez — puis modifiez les paramètres pour trouver le régime où la profondeur cesse d'être dispensable.\n", - "\n", - "Indices :\n", - "- `# Étape 1` : lancer la cellule telle quelle et comparer les deux accuracies affichées.\n", - "- `# Étape 2` : réduire la largeur à 4 — remplacer `dims=(2, 8, 1)` par `(2, 4, 1)` et `(2, 4, 4, 1)`. La profondeur compense-t-elle l'étroitesse ?\n", - "- `# Étape 3` : augmenter le bruit (`noise=0.25`) — quel réseau résiste mieux au chevauchement des classes ?\n", - "- Lecture attendue : contre l'intuition, la profondeur peut **faire perdre** à budget égal — comptez les sigmoïdes que le gradient doit traverser pour revenir à la première couche. Puis doublez les itérations : que devient la comparaison ?\n" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "id": "440bb65c", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:18.000328Z", - "iopub.status.busy": "2026-08-23T01:26:17.999329Z", - "iopub.status.idle": "2026-08-23T01:26:18.960020Z", - "shell.execute_reply": "2026-08-23T01:26:18.959501Z" - }, - "papermill": { - "duration": 0.971733, - "end_time": "2026-08-23T01:26:18.960535+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:17.988802+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1 couche cachee (2-8-1) : loss train = 0.0784 | accuracy test = 0.983\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "2 couches cachees (2-8-8-1) : loss train = 0.3773 | accuracy test = 0.783\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Mini-cas make_circles : le meme backward, generalise a un nombre quelconque de couches\n", - "from sklearn.datasets import make_circles\n", - "\n", - "def forward_gen(Xd, layers):\n", - " \"\"\"Passe avant generique : A[0] = X, A[k+1] = sigmoid(A[k] @ W_k + b_k), A[-1] = prediction.\"\"\"\n", - " A = [Xd]\n", - " for W, b in layers:\n", - " A.append(sigmoid(A[-1] @ W + b))\n", - " return A\n", - "\n", - "def backward_gen(layers, y, A):\n", - " \"\"\"Passe arriere generique : le meme motif delta -> gradients -> delta, en boucle.\"\"\"\n", - " N = y.shape[0]\n", - " grads = [None] * len(layers)\n", - " delta = (A[-1] - y.reshape(-1, 1)) / N\n", - " for k in range(len(layers) - 1, -1, -1):\n", - " grads[k] = (A[k].T @ delta, delta.sum(axis=0))\n", - " if k > 0:\n", - " delta = (delta @ layers[k][0].T) * A[k] * (1.0 - A[k])\n", - " return grads\n", - "\n", - "def init_layers(dims, seed=42):\n", - " \"\"\"Xavier couche par couche : W_k ~ N(0, 1/fan_in), biais a zero.\"\"\"\n", - " rng = np.random.default_rng(seed)\n", - " return [\n", - " (rng.normal(0.0, np.sqrt(1.0 / fan_in), size=(fan_in, fan_out)), np.zeros(fan_out))\n", - " for fan_in, fan_out in zip(dims[:-1], dims[1:])\n", - " ]\n", - "\n", - "def train_gen(layers, Xd, yd, lr=0.5, iters=4000):\n", - " for _ in range(iters):\n", - " A = forward_gen(Xd, layers)\n", - " grads = backward_gen(layers, yd, A)\n", - " layers = [(W - lr * dW, b - lr * db) for (W, b), (dW, db) in zip(layers, grads)]\n", - " return layers\n", - "\n", - "def gen_fwd(layers, Xd):\n", - " return forward_gen(Xd, layers)[-1], None\n", - "\n", - "Xc, yc = make_circles(n_samples=400, noise=0.15, factor=0.4, random_state=42)\n", - "Xc_tr, Xc_te, yc_tr, yc_te = train_test_split(Xc, yc, test_size=0.3, random_state=42, stratify=yc)\n", - "\n", - "fig, axes = plt.subplots(1, 2, figsize=(11, 4.4))\n", - "for ax, (nom, dims) in zip(axes, [(\"1 couche cachee (2-8-1)\", (2, 8, 1)),\n", - " (\"2 couches cachees (2-8-8-1)\", (2, 8, 8, 1))]):\n", - " Lc = train_gen(init_layers(dims, seed=42), Xc_tr, yc_tr, lr=0.5, iters=4000)\n", - " acc = ((gen_fwd(Lc, Xc_te)[0] > 0.5).astype(int).ravel() == yc_te).mean()\n", - " loss_tr = bce_loss(gen_fwd(Lc, Xc_tr)[0], yc_tr)\n", - " print(f\"{nom:28s} : loss train = {loss_tr:.4f} | accuracy test = {acc:.3f}\")\n", - " plot_boundary(ax, Lc, Xc_te, yc_te, f\"{nom} -- acc test = {acc:.2f}\", fwd=gen_fwd)\n", - "fig.suptitle(\"make_circles : la profondeur achete-t-elle quelque chose ?\")\n", - "fig.tight_layout(); plt.show()\n" - ] - }, - { - "cell_type": "markdown", - "id": "3bfa38b2", - "metadata": { - "papermill": { - "duration": 0.010694, - "end_time": "2026-08-23T01:26:18.982149+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:18.971455+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "### Lecture du résultat\n", - "\n", - "Surprise mesurée : à budget égal (4 000 itérations), le réseau **profond** fait moins bien que le shallow — accuracy 78 % contre 98 %, loss 0.38 contre 0.08. Ce n'est pas un plafond d'expressivité (deux couches savent représenter tout ce qu'une sait représenter), c'est un coût d'**optimisation** : chaque couche de sigmoïde traversée multiplie le gradient qui remonte par un facteur $\\sigma'(z) = a(1-a) \\le 0.25$ — le gradient s'amenuise couche après couche (*vanishing gradient*) et les premières couches apprennent au ralenti. Doublez le budget (exercice 2) : le profond rattrape son retard et finit par prendre l'avantage en loss finale. La profondeur achète de l'expressivité mais la paie en vitesse de convergence — ce constat est précisément ce qui a motivé les activations ReLU puis les normalisations, au programme de la suite de la sous-série.\n" - ] - }, - { - "cell_type": "markdown", - "id": "7399ff8a", - "metadata": { - "papermill": { - "duration": 0.010715, - "end_time": "2026-08-23T01:26:19.003873+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:18.993158+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## 8. Capstone : la backward vue comme composition de Jacobiennes\n", - "\n", - "La chain rule tensorielle se réécrit de façon compacte : chaque étape du réseau est une fonction, sa dérivée est une **matrice Jacobienne**, et la rétropropagation est le produit de ces matrices — appliqué **dans l'ordre inverse** et sans jamais les matérialiser en entier (on ne transporte que des *vecteurs* à travers elles : des produits vecteur-Jacobienne). C'est ce qui fait que backpropager coûte à peine plus cher qu'une passe avant, quel que soit le nombre de paramètres.\n", - "\n", - "### Exercice 3 (capstone) : la backward comme composition Jacobienne\n", - "\n", - "Pour le réseau $x \\to h = \\sigma(W_1^\\top x + b_1) \\to \\hat{y} = \\sigma(w_2^\\top h + b_2)$ (dimensions 2 → 4 → 1, convention vecteur colonne — la transposée de la convention matricielle en batch des sections précédentes) et **un seul** exemple $(x, y)$ :\n", - "\n", - "1. Écrivez les quatre Jacobiennes : $J_1 = \\partial z_1 / \\partial x$ (4×2), $J_{\\sigma_1} = \\partial a_1 / \\partial z_1$ (4×4 diagonale), $J_2 = \\partial z_2 / \\partial a_1$ (1×4), $J_{\\sigma_2} = \\partial \\hat{y} / \\partial z_2$ (scalaire).\n", - "2. Calculez $\\partial L / \\partial x$ en multipliant **dans l'ordre inverse**, en ne transportant que des vecteurs : une simple ligne NumPy.\n", - "3. Vérifiez chaque composante par différence finie sur $x$ (réutilisez le motif de la section 2) — écart attendu < 1e-7.\n", - "\n", - "Indices :\n", - "- `# Étape 1` : $J_1 = W_1^\\top$, $J_{\\sigma_1} = \\mathrm{diag}(a_1 \\odot (1 - a_1))$, $J_2 = w_2^\\top$.\n", - "- `# Étape 2` : en NumPy, inutile de matérialiser la diagonale — multiplier terme à terme par $a_1 \\odot (1 - a_1)$ suffit.\n", - "- `# Étape 3` : $(L(x + \\varepsilon e_i) - L(x - \\varepsilon e_i)) / (2 \\varepsilon)$ pour $i = 1, 2$, avec $\\varepsilon = 10^{-5}$.\n", - "- Question bonus : pourquoi backpropager coûte-t-il à peu près une passe avant, quel que soit le nombre de paramètres ? (indice : on ne matérialise jamais les Jacobiennes complètes, seulement des produits vecteur-Jacobienne.)\n" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "id": "56cdc793", - "metadata": { - "execution": { - "iopub.execute_input": "2026-08-23T01:26:19.027501Z", - "iopub.status.busy": "2026-08-23T01:26:19.027501Z", - "iopub.status.idle": "2026-08-23T01:26:19.032996Z", - "shell.execute_reply": "2026-08-23T01:26:19.031988Z" - }, - "papermill": { - "duration": 0.0176, - "end_time": "2026-08-23T01:26:19.032996+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:19.015396+00:00", - "status": "completed" - }, - "tags": [] - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Exercice a completer : dL/dx via produits vecteur-Jacobien + verification numerique < 1e-7\n" - ] - } - ], - "source": [ - "# Exercice 3 (capstone) : gradient dL/dx comme composition de Jacobienne (reseau 2 -> 4 -> 1, 1 exemple)\n", - "x_ex = np.array([[0.8, -0.6]]) # un seul exemple, en ligne comme dans le reste du notebook\n", - "y_ex = 1.0\n", - "# TODO etudiant : tirer W1 (2x4), b1 (4,), w2 (4,1), b2 (1,) avec np.random.default_rng(0), passe avant\n", - "# TODO etudiant : ecrire dL/dx comme produits vecteur-Jacobien, dans l'ORDRE INVERSE (reverse-mode)\n", - "dL_dx = None # TODO etudiant : vecteur (1, 2), a verifier composante par composante par difference finie\n", - "print(\"Exercice a completer : dL/dx via produits vecteur-Jacobien + verification numerique < 1e-7\")\n" - ] - }, - { - "cell_type": "markdown", - "id": "2e91c631", - "metadata": { - "papermill": { - "duration": 0.011005, - "end_time": "2026-08-23T01:26:19.055609+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:19.044604+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## Synthèse et transition\n", - "\n", - "Trois leçons à retenir :\n", - "\n", - "1. **La rétropropagation est la chain rule, rien de plus.** Un $\\delta$ remonte de la sortie vers l'entrée ; chaque couche l'utilise pour calculer ses gradients avant de le propager à la couche précédente. Le couple BCE + sigmoïde rend le tout élégant : $\\delta_2 = \\hat{y} - y$, l'erreur de prédiction elle-même.\n", - "2. **La vérification numérique du gradient est la preuve que le code est juste.** Écart analytique/numérique < 1e-7 par différence finie : un filet de sécurité à déployer chaque fois qu'on écrit une dérivée à la main — ici comme dans n'importe quel moteur différentiable.\n", - "3. **PyTorch exécute exactement ce qu'on a écrit** — mêmes trajectoires, mêmes poids à convergence. Pas de magie dans `backward()` : c'est ce qui le rend fiable. Et l'initialisation n'est pas un détail : zéros = symétrie gelée à $\\ln 2$ ; Xavier = un réseau qui apprend.\n", - "\n", - "Transition : ce réseau a 33 paramètres et tient dans une cellule. Le notebook suivant (`3.2-Entraînement-GPU`) passe aux réseaux plus profonds, aux mini-batchs et au GPU — la mécanique ne change pas, seule l'échelle change.\n" - ] - }, - { - "cell_type": "markdown", - "id": "1800d3dc", - "metadata": { - "papermill": { - "duration": 0.010109, - "end_time": "2026-08-23T01:26:19.076226+00:00", - "exception": false, - "start_time": "2026-08-23T01:26:19.066117+00:00", - "status": "completed" - }, - "tags": [] - }, - "source": [ - "## References\n", - "\n", - "1. Rumelhart, D., Hinton, G. & Williams, R. (1986). *Learning representations by back-propagating errors*. Nature 323, 533–536. — L'article fondateur : la rétropropagation appliquée aux réseaux multicouches.\n", - "2. Glorot, X. & Bengio, Y. (2010). *Understanding the difficulty of training deep feedforward neural networks*. AISTATS 2010. — Pourquoi l'initialisation compte : variance $1/\\mathrm{fan\\_in}$ et rupture de symétrie.\n", - "3. Nielsen, M. *Neural Networks and Deep Learning*, chapitre 2 — une dérivation pas à pas de la rétropropagation, avec le point de vue Jacobien.\n", - "4. Baydin, A., Pearlmutter, B., Radul, A. & Siskind, J. (2018). *Automatic Differentiation in Machine Learning: a Survey*. JMLR 18. — Forward vs reverse mode : ce qu'autograd fait en silence.\n" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.11.15" - }, - "papermill": { - "default_parameters": {}, - "duration": 24.844349, - "end_time": "2026-08-23T01:26:20.284598+00:00", - "environment_variables": {}, - "exception": null, - "input_path": "MyIA.AI.Notebooks/ML/DataScienceWithAgents/03-DeepLearning/3.1-Retropropagation-From-Scratch.ipynb", - "output_path": "C:/Users/jsboi/AppData/Local/Temp/claude/c--dev-CoursIA-2/10ac7e0c-f66d-482a-8a02-150b7a45c479/scratchpad/3.1-executed.ipynb", - "parameters": {}, - "start_time": "2026-08-23T01:25:55.440249+00:00", - "version": "2.7.0" - } - }, - "nbformat": 4, - "nbformat_minor": 5 -}