2022/07/23 released
- Python 3.8.13
- CUDA 11.3
- Pytorch 1.11
# 使用Anaconda建立名為training的虛擬環境
$ conda create --name training python=3.8.13
# 啟動名為training的虛擬環境
$ conda activate training
$ git clone https://github.com/Offliners/RobotLab-MLDL-Training-2022.git
$ cd RobotLab-MLDL-Training-2022
# 安裝此專案所需的所有函式庫
$ pip install -r requirements.txt
# 前往想要訓練的範例
$ cd tutorial-x
# 完成訓練後即可關閉虛擬環境
$ conda deactivate
# 若之後確定不會使用的話,可以移除虛擬環境
$ conda env remove -n training
Press F12
,and enter this code in console,then press enter
function ClickConnect(){
console.log("Connnect Clicked - Start");
document.querySelector("#top-toolbar > colab-connect-button").shadowRoot.querySelector("#connect").click();
console.log("Connnect Clicked - End");
};
setInterval(ClickConnect, 60000)
insert code cell at the bottom
from google.colab import files
files.download("output_file.csv") # "output_file.csv" must be your output file name
insert code cell to check which GPU is assigned
!nvidia-smi
Copy the URL to https://nbviewer.jupyter.org/
透過相關係數的分析,選出重要的特徵來訓練模型做Covid19 Cases Prediction
Dataset Resource : Covid19 Cases
Links : [README] [Jupyter Notebook] [Google Colab]
透過Teacher-Student的架構,訓練出3個學生模型,並使用Ensemble與Test Time Augmentation的技術來完成Food Classification
Dataset Resource : Food11
Links : [README] [Jupyter Notebook] [Google Colab]
使用resnet18當backbone,並用UNet來完成Synthetic Image Segmentation
Links : [README] [Jupyter Notebook] [Google Colab]
使用SNGAN來完成Anime Face Generation
Dataset Resource : Crypko
Links : [README] [Jupyter Notebook] [Google Colab]
使用PPO策略來訓練模型完成Super Mario任務
World/Stage | 1 | 2 | 3 | 4 |
---|---|---|---|---|
1 | ||||
2 | ||||
3 | ||||
4 | ||||
5 | ||||
6 | ||||
7 | ||||
8 |
Environment : gym-super-mario-bros
Links : [README] [Jupyter Notebook] [Google Colab]
- Pytorch official tutorials
- NTU Applied Deep Learning 2019 Spring (CSIE5431, by Prof. Yun-Nung Chen)
- NTU Machine Learning 2021 Spring (EE5184, by Prof. Hung-yi Lee)
- NTU Machine Learning 2022 Spring (EE5184, by Prof. Hung-yi Lee)
- MIT Deep Learning and Artificial Intelligence Lectures
- UC Berkeley Deep Reinforcement Learning 2021 Fall (CS285, by Prof. Sergey Levine)
- UNet/FCN PyTorch
- Super-mario-bros-PPO-pytorch
- mario-ai-challenge