| Completed | # | Topic | Task | Public Baseline | Private Baseline |
|---|---|---|---|---|---|
| ✅ | HW 1 | Regression | 预测 COVID-19 新增病例 | Boss | Strong |
| ✅ | HW 2 | Classification | Phoneme 分类 | Strong | Strong |
| ✅ | HW 3 | CNN | 食物分类 | Boss | Boss |
| ✅ | HW 4 | Self-Attention | Speaker 识别 | Boss | Boss |
| HW 5 | Transformer | ||||
| HW 6 | Generative Model | ||||
| HW 7 | BERT | ||||
| HW 8 | Auto-encoder | ||||
| HW 9 | Explainable AI | ||||
| HW 10 | Attack | ||||
| HW 11 | Adaptation | ||||
| HW 12 | Reinforcement Learning | ||||
| HW 13 | Network Compression | ||||
| HW 14 | Life-long Learning | ||||
| HW 15 | Meta Learning |
https://github.com/hsushuai/ml2023spring-hw.git
cd ml2023spring
pip install -r requirements.txt在 releases 中下载对应作业的数据文件,或使用 wget 下载,以 HW 1 为例:
wget https://github.com/hsushuai/ml2023spring-hw/releases/download/dataset/ml2023spring-hw1.zip
unzip ml2023spring-hw1.zip以 HW 1 为例:
python main.py hw1运行命令必须指定需要运行的作业,可以是从 hw1 一直到 hw15。
此外,你还可以在命令中添加可选参数,参数名称为对应作业中对应的 configs。以 HW 1 为例,可以在命令中任意设置 hw1-config 中的参数,比如:
python main.py hw1 --max_epochs 3000 --data_dir /mnt/data --output_dir /mnt/output