[Example] Add VAE example #573
Conversation
|
Thanks for your contribution! |
|
✅ This PR's description meets the template requirements! |
|
感谢大佬提PR，可以合并一下最新develop分支的代码，然后案例可以用hydra改造一下，参考这个案例：https://github.com/PaddlePaddle/PaddleScience/tree/develop/examples/bracket |
|
@HydrogenSulfate 不好意思，打扰了，AISTUDIO原文用到了 原文地址 ：https://aistudio.baidu.com/projectdetail/5541961 # julia 依赖
os.environ['JULIA_DEPOT_PATH'] = '/home/aistudio/opt/julia_package'
# pip 依赖
sys.path.append('/home/aistudio/opt/external-libraries')
# julieries
from julia.api import Julia
jl = Julia(compiled_modules=False,runtime="/home/aistudio/opt/julia-1.8.5/bin/julia")
# import julia
from julia import Main |
测试文件test.ipynb只是一个测试功能是否正常的文件，本地测试使用，可以不用提交到PR里。 |
可以将上述路径配置写到hydra yaml文件里，然后通过cfg.JULIA.xxx访问即可。然后在代码里加几行检测路径是否存在的代码，提醒用户正确设置julia路径。此外在案例文档中，可以加上Julia配置这一章节，里面包含了对配置文件的julia路径设置 |
| # def train_mse_func( | ||
| # output_dict: Dict[str, "paddle.Tensor"], label_dict: Dict[str, "pgl.Graph"], *args | ||
| # ) -> paddle.Tensor: | ||
| # return F.mse_loss(output_dict["pred"], label_dict["label"].y) |
examples/RegAE/train_new.py
Outdated
| latent_dim, hidden_dim = 100, 100 | ||
| # set model | ||
| model = ppsci.arch.AutoEncoder( | ||
| input_dim=10000, | ||
| latent_dim=latent_dim, | ||
| hidden_dim=hidden_dim, | ||
| ) |
There was a problem hiding this comment.
AutoEncoder的初始化超参数移动至RegAE.yaml配置文件里
examples/RegAE/train_new.py
Outdated
| # pretrained_model_path="./output_AMGNet/checkpoints/latest" | ||
| ) | ||
| # train model | ||
| solver.train() |
| class AutoEncoder(base.Arch): | ||
| def __init__(self, input_dim, latent_dim, hidden_dim): | ||
| super(AutoEncoder, self).__init__() | ||
|
|
||
| # encoder | ||
| self._encoder_linear = nn.Sequential( | ||
| nn.Linear(input_dim, hidden_dim), | ||
| nn.Tanh(), | ||
| ) | ||
| self._encoder_mu = nn.Linear(hidden_dim, latent_dim) | ||
| self._encoder_log_sigma = nn.Linear(hidden_dim, latent_dim) | ||
|
|
||
| self._decoder = nn.Sequential( | ||
| nn.Linear(latent_dim, hidden_dim), | ||
| nn.Tanh(), | ||
| nn.Linear(hidden_dim, input_dim), | ||
| ) |
There was a problem hiding this comment.
base.Arch的子类需要做成符号化的形式，即需要具有 input_keys、output_keys两个属性
| def __getitem__(self, idx): | ||
| if self.data_type == "train": | ||
| return self.train_data[idx] | ||
| else: | ||
| return self.test_data[idx] |
There was a problem hiding this comment.
dataset需要返回input、label、weight三个数据字典
| # 计算mu，log_sigma与 N(0,1)分布的差距 | ||
| base = paddle.exp(2. * log_sigma) + paddle.pow(mu, 2) - 1. - 2. * log_sigma | ||
| loss = 0.5 * paddle.sum(base) / mu.shape[0] | ||
| return loss |
| def __init__(self): | ||
| super().__init__(None, None) | ||
|
|
||
| def forward(self, mu, log_sigma): |
There was a problem hiding this comment.
所有loss子类的forward写法统一，接受output_dict, label_dict, weight_dict三个变量，然后计算loss并以一个标量Tensor的形式返回。可以将N(0,1)的数据作为标签数据，从label_dict传入
|
closed for #660 |
PR types
New features
PR changes
Others
Describe
目前约完成 1/2 还在精度对齐ing