README.md

@@ -2,209 +2,40 @@
 
 MSO 3.4 Apprentissage Automatique
 
-# 
+# Objectif du TD
 
-In this hands-on project, we will first implement a simple RL algorithm and apply it to solve the CartPole-v1 environment. Once we become familiar with the basic workflow, we will learn to use various tools for machine learning model training, monitoring, and sharing, by applying these tools to train a robotic arm.
+Dans ce projet pratique, nous commencerons par mettre en œuvre un algorithme RL simple et nous l'appliquerons pour résoudre l'environnement CartPole-v1. Une fois que nous nous serons familiarisés avec le flux de travail de base, nous apprendrons à utiliser divers outils pour la formation, le contrôle et le partage de modèles d'apprentissage automatique, en appliquant ces outils à la formation d'un bras robotique.
 
-## To be handed in
 
-This work must be done individually. The expected output is a repository named `hands-on-rl` on https://gitlab.ec-lyon.fr. 
+## Détails des fichiers python:
+# reinforce_cartpole.py: implemention de l'algorithme de RL en utilisant Pytorch
 
-We assume that `git` is installed, and that you are familiar with the basic `git` commands. (Optionnaly, you can use GitHub Desktop.)
-We also assume that you have access to the [ECL GitLab](https://gitlab.ec-lyon.fr/). If necessary, please consult [this tutorial](https://gitlab.ec-lyon.fr/edelland/inf_tc2/-/blob/main/Tutoriel_gitlab/tutoriel_gitlab.md).
+Classe PolicyNetwork : Définit un réseau neuronal simple pour servir de politique. Il s'agit d'un réseau neuronal feedforward avec une couche cachée et une fonction d'activation ReLU.
 
-Your repository must contain a `README.md` file that explains **briefly** the successive steps of the project. It must be private, so you need to add your teacher as "developer" member.
+Fonction compute_returns : Calcule les rendements actualisés pour un épisode donné. Les rendements actualisés sont un composant crucial de l'algorithme REINFORCE. Elle calcule la somme des récompenses, actualisées par un facteur (gamma) pour les récompenses futures, puis normalise ces rendements.
 
-Throughout the subject, you will find a 🛠 symbol indicating that a specific production is expected.
+Fonction reinforce : Implémente l'algorithme REINFORCE. Elle itère sur un nombre spécifié d'épisodes, interagit avec l'environnement, collecte des expériences (états, actions, récompenses), calcule les rendements et met à jour le réseau de politique en utilisant les rendements calculés et les log-probabilités.
 
-The last commit is due before 11:59 pm on March 5, 2024. Subsequent commits will not be considered.
+Bloc principal : Entraîne le réseau de politique en utilisant l'algorithme REINFORCE, puis trace la récompense totale obtenue à chaque épisode pendant l'entraînement. Il enregistre le tracé généré dans un répertoire nommé 'plots'.
 
-> ⚠️ **Warning**
-> Ensure that you only commit the files that are requested. For example, your directory should not contain the generated `.zip` files, nor the `runs` folder... At the end, your repository must contain one `README.md`, three python scripts, and optionally image files for the plots.
+Traçage : Trace la récompense totale obtenue à chaque épisode pendant l'entraînement.
 
-## Before you start
+Enregistrement du tracé : Enregistre le tracé généré sous forme de fichier PNG dans le répertoire 'plots'.
 
-Make sure you know the basics of Reinforcement Learning. In case of need, you can refer to the [introduction of the Hugging Face RL course](https://huggingface.co/blog/deep-rl-intro).
+# a2c_sb3_cartpole.py: 
+implemention de l'algorithme de RL en utilisant stable_baselines3 et A2C algorithm.
 
-## Introduction to Gym
+# a2c_sb3_panda_reach.py: 
+implemention du modele précedent avec l'environement PandaReachJointsDense-v2.
 
-[Gym](https://gymnasium.farama.org/) is a framework for developing and evaluating reinforcement learning environments. It offers various environments, including classic control and toy text scenarios, to test RL algorithms.
 
-### Installation
 
-We recommend to use Python virtual environnements to install the required modules : https://docs.python.org/3/library/venv.html
 
-First, install Pytorch : https://pytorch.org/get-started/locally.
+## Lien vers Hugging face: 
+https://huggingface.co/MohamedKhalil
 
-Then install the following modules :
+L'utilisation de la fonction push_to_hub a été sans succès. Donc le partage du modèle sur huggingface n'a pas abouti.
 
 
-```sh
-pip install gym==0.26.2
-```
 
-Install also pyglet for the rendering.
 
-```sh
-pip install pyglet==2.0.10
-```
-
-If needed 
-
-```sh
-pip install pygame==2.5.2
-```
-
-```sh
-pip install PyQt5
-```
-
-
-### Usage
-
-Here is an example of how to use Gym to solve the `CartPole-v1` environment [Documentation](https://gymnasium.farama.org/environments/classic_control/cart_pole/):
-
-```python
-import gym
-
-# Create the environment
-env = gym.make("CartPole-v1", render_mode="human")
-
-# Reset the environment and get the initial observation
-observation = env.reset()
-
-for _ in range(100):
-    # Select a random action from the action space
-    action = env.action_space.sample()
-    # Apply the action to the environment
-    # Returns next observation, reward, done signal (indicating
-    # if the episode has ended), and an additional info dictionary
-    observation, reward, terminated, truncated, info = env.step(action)
-    # Render the environment to visualize the agent's behavior
-    env.render()
-    if terminated: 
-        # Terminated before max step
-        break
-
-env.close()
-```
-
-## REINFORCE
-
-The REINFORCE algorithm (also known as Vanilla Policy Gradient) is a policy gradient method that optimizes the policy directly using gradient descent. The following is the pseudocode of the REINFORCE algorithm:
-
-```txt
-Setup the CartPole environment
-Setup the agent as a simple neural network with:
-    - One fully connected layer with 128 units and ReLU activation followed by a dropout layer
-    - One fully connected layer followed by softmax activation
-Repeat 500 times:
-    Reset the environment
-    Reset the buffer
-    Repeat until the end of the episode:
-        Compute action probabilities 
-        Sample the action based on the probabilities and store its probability in the buffer 
-        Step the environment with the action
-        Compute and store in the buffer the return using gamma=0.99 
-    Normalize the return
-    Compute the policy loss as -sum(log(prob) * return)
-    Update the policy using an Adam optimizer and a learning rate of 5e-3
-```
-
-To learn more about REINFORCE, you can refer to [this unit](https://huggingface.co/learn/deep-rl-course/unit4/introduction).
-
-> 🛠 **To be handed in**
-> Use PyTorch to implement REINFORCE and solve the CartPole environement. Share the code in `reinforce_cartpole.py`, and share a plot showing the total reward accross episodes in the `README.md`.
-
-## Familiarization with a complete RL pipeline: Application to training a robotic arm
-
-In this section, you will use the Stable-Baselines3 package to train a robotic arm using RL. You'll get familiar with several widely-used tools for training, monitoring and sharing machine learning models.
-
-### Get familiar with Stable-Baselines3
-
-Stable-Baselines3 (SB3) is a high-level RL library that provides various algorithms and integrated tools to easily train and test reinforcement learning models.
-
-#### Installation
-
-```sh
-pip install stable-baselines3
-pip install moviepy
-```
-
-#### Usage
-
-Use the [Stable-Baselines3 documentation](https://stable-baselines3.readthedocs.io/en/master/) to implement the code to solve the CartPole environment with the Advantage Actor-Critic (A2C) algorithm.
-
-
-> 🛠 **To be handed in**
-> Store the code in `a2c_sb3_cartpole.py`. Unless otherwise stated, you'll work upon this file for the next sections.
-
-### Get familiar with Hugging Face Hub
-
-Hugging Face Hub is a platform for easy sharing and versioning of trained machine learning models. With Hugging Face Hub, you can quickly and easily share your models with others and make them usable through the API. For example, see the trained A2C agent for CartPole: https://huggingface.co/sb3/a2c-CartPole-v1. Hugging Face Hub provides an API to download and upload SB3 models.
-
-#### Installation of `huggingface_sb3`
-
-```sh
-pip install huggingface-sb3==2.3.1
-```
-
-#### Upload the model on the Hub
-
-Follow the [Hugging Face Hub documentation](https://huggingface.co/docs/hub/stable-baselines3) to upload the previously learned model to the Hub.
-
-> 🛠 **To be handed in**
-> Link the trained model in the `README.md` file.
-
-> 📝 **Note**
->  [RL-Zoo3](https://stable-baselines3.readthedocs.io/en/master/guide/rl_zoo.html) provides more advanced features to save hyperparameters, generate renderings and metrics. Feel free to try them.
-
-### Get familiar with Weights & Biases
-
-Weights & Biases (W&B) is a tool for machine learning experiment management. With W&B, you can track and compare your experiments, visualize your model training and performance.
-
-#### Installation
-
-You'll need to install both `wand` and `tensorboar`.
-
-```shell
-pip install wandb tensorboard
-```
-
-Use the documentation of [Stable-Baselines3](https://stable-baselines3.readthedocs.io/en/master/) and [Weights & Biases](https://docs.wandb.ai/guides/integrations/stable-baselines-3) to track the CartPole training. Make the run public.
-
-🛠 Share the link of the wandb run in the `README.md` file.
-
-> ⚠️ **Warning**
-> Make sure to make the run public!
-
-### Full workflow with panda-gym
-
-[Panda-gym](https://github.com/qgallouedec/panda-gym) is a collection of environments for robotic simulation and control. It provides a range of challenges for training robotic agents in a simulated environment. In this section, you will get familiar with one of the environments provided by panda-gym, the `PandaReachJointsDense-v3`. The objective is to learn how to reach any point in 3D space by directly controlling the robot's articulations.
-
-#### Installation
-
-```shell
-pip install panda-gym==3.0.7
-```
-
-#### Train, track, and share
-
-Use the Stable-Baselines3 package to train A2C model on the `PandaReachJointsDense-v2` environment. 500k timesteps should be enough. Track the environment with Weights & Biases. Once the training is over, upload the trained model on the Hub.
-
-> 🛠 **To be handed in**
-> Share all the code in `a2c_sb3_panda_reach.py`. Share the link of the wandb run and the trained model in the `README.md` file.
-
-## Contribute
-
-This tutorial may contain errors, inaccuracies, typos or areas for improvement. Feel free to contribute to its improvement by opening an issue.
-
-## Author
-
-Quentin Gallouédec
-
-Updates by Léo Schneider, Emmanuel Dellandréa
-
-## License
-
-MIT

a2c_sb3_cartpole.py

+import gym
+import numpy as np
+import matplotlib.pyplot as plt
+from stable_baselines3 import A2C
+from transformers.hf_api import push_to_hub
+
+
+# Create the CartPole environment
+env = gym.make("CartPole-v1")
+
+# Create the A2C model
+model = A2C("MlpPolicy", env, verbose=1)
+
+# Train the model
+model.learn(total_timesteps=10000)
+
+# Get environment for rendering
+vec_env = model.get_env()
+
+# Plotting the rewards
+episode_rewards = []
+obs = vec_env.reset()
+for _ in range(500):
+    action, _state = model.predict(obs, deterministic=True)
+    obs, reward, done, info = vec_env.step(action)
+    # done = terminated or truncated
+    episode_rewards.append(reward)
+    if done:
+        obs = vec_env.reset()
+
+# Plotting the rewards
+plt.plot(np.arange(len(episode_rewards)), episode_rewards)
+plt.xlabel('Time Steps')
+plt.ylabel('Reward')
+plt.title('Reward Across Time Steps')
+plt.show()
+
+
+#save the model
+model.save("cartpole_model_a2c_sb3")
+
+
+push_to_hub(
+    repo_id="MohamedKhalil/Hands-on-rl",
+    filename="cartpole_model_a2c_sb3",
+    commit_message="First push",
+)
\ No newline at end of file

a2c_sb3_panda_reach.py

+import gym
+import wandb
+import panda_gym
+import stable_baselines3
+from stable_baselines3 import A2C
+import matplotlib.pyplot as plt 
+
+# start a new wandb run to track this script
+wandb.init(project='Model2',   
+    config={
+    "Algorithm": "A2C",
+    "env": "CartPole-v1",
+    "episodes": 500,
+    })
+#define the environment
+env = gym.make('PandaReachJointsDense-v2')
+model = stable_baselines3.A2C('MultiInputPolicy', env, verbose=1)
+
+scores=[]
+for i in range(100):
+    model.learn(total_timesteps=1000)
+    obs = env.reset()
+    rewards=0
+    while True:
+        action, _states = model.predict(obs)
+        obs, reward, done, info = env.step(action)
+        rewards+=reward
+        env.render()
+        if done:
+            break
+    scores.append(rewards)
+    wandb.log({"rewards" : scores})
+        
+
+wandb.join()
+model.save("PandaReach")
+
+plt.plot(scores)
+plt.xlabel("Episode")
+plt.ylabel("Total Reward")
+plt.show()

reinforce_cartpole.py

+import gym
+import torch
+import torch.nn as nn
+import torch.optim as optim
+import numpy as np
+import matplotlib.pyplot as plt
+import os
+
+# Define a simple neural network as the policy
+class PolicyNetwork(nn.Module):
+    def __init__(self, input_size, output_size, hidden_size=128):
+        super(PolicyNetwork, self).__init__()
+        self.fc1 = nn.Linear(input_size, hidden_size)
+        self.relu = nn.ReLU()
+        self.fc2 = nn.Linear(hidden_size, output_size)
+        self.softmax = nn.Softmax(dim=-1)
+
+    def forward(self, x):
+        x = self.relu(self.fc1(x))
+        x = self.fc2(x)
+        return self.softmax(x)
+
+# Function to compute discounted returns
+def compute_returns(rewards, gamma=0.99):
+    returns = []
+    R = 0
+    for r in reversed(rewards):
+        R = r + gamma * R
+        returns.insert(0, R)
+    returns = torch.tensor(returns)
+    returns = (returns - returns.mean()) / (returns.std() + 1e-8)  # Normalize
+    return returns
+
+# Function to train the policy network using REINFORCE algorithm
+def reinforce(env_name="CartPole-v1", num_episodes=500, gamma=0.99, learning_rate=5e-3):
+    # Initialize environment and policy network
+    env = gym.make(env_name)
+    policy = PolicyNetwork(env.observation_space.shape[0], env.action_space.n)
+    optimizer = optim.Adam(policy.parameters(), lr=learning_rate)
+    
+    episode_rewards = []
+    
+    for episode in range(num_episodes):
+        episode_reward = 0
+        log_probs = []
+        rewards = []
+        
+        state = env.reset()
+        done = False
+
+        while not done:
+            state_array = state[0] if isinstance(state, tuple) else state  # Extract array representing the state
+            state_tensor = torch.tensor(state_array, dtype=torch.float32).unsqueeze(0)  # Convert to tensor
+            action_probs = policy(state_tensor)
+            action = torch.multinomial(action_probs, num_samples=1).item()
+            log_probs.append(torch.log(action_probs[0][action]))  # Access the first element of the tensor
+
+            # Take action in the environment
+            step_output = env.step(action)
+            next_state, reward, terminated,truncated, info = step_output  # Unpack step output
+            done = truncated or terminated
+
+            rewards.append(reward)  # Collect rewards
+
+            # Update the state for the next iteration
+            state = next_state
+            
+            episode_reward += reward  # Accumulate rewards
+
+        # Compute discounted returns
+        returns = compute_returns(rewards, gamma)
+        
+        # Compute loss and update policy
+        policy_loss = -torch.sum(torch.stack(log_probs) * returns)
+        optimizer.zero_grad()
+        policy_loss.backward()
+        optimizer.step()
+        
+        episode_rewards.append(episode_reward)
+        print(f"Episode {episode + 1}/{num_episodes}, Reward: {episode_reward}")
+        
+    env.close()
+    
+    return episode_rewards
+
+
+
+if __name__ == "__main__":
+    # Train the policy network
+    episode_rewards = reinforce()
+
+    # Plotting the total reward across episodes
+    plt.plot(np.arange(len(episode_rewards)), episode_rewards)
+    plt.xlabel('Episode')
+    plt.ylabel('Total Reward')
+    plt.title('Total Reward Across Episodes')
+
+    # Save the plot in a file named 'plots'
+    plots_dir = 'plots'
+    if not os.path.exists(plots_dir):
+        os.makedirs(plots_dir)
+    plot_file = os.path.join(plots_dir, 'reward_plot1.png')
+    plt.savefig(plot_file)
+
+    plt.show()