Conditional Graph Generation Issue
yikang613 opened this issue · comments
Hi, I met a question about the regressor model. I think the model does not correctly performed the conditional generation.
The error for this command python3 src/guidance/train_qm9_regressor.py +experiment=regressor_model.yaml happens at the dimension of graph labels.
Here is the error.
target: tensor([], size=(512, 0))
Shape of pred.y: torch.Size([512, 1])
Shape of target: torch.Size([512, 0])
Can you please debug it?
I have the same problem, can you describe the solution in more detail?
Hello, can you make sure that you reprocessed the dataset, as specified in the README of the branch? Thank you
…
On 27 Sep 2023, at 04:07, yikang613 @.***> wrote: Hi, I met a question about the regressor model. I think the model does not correctly performed the conditional generation. The error for this command python3 src/guidance/train_qm9_regressor.py +experiment=regressor_model.yaml happens at the dimension of graph labels. Here is the error. target: tensor([], size=(512, 0)) Shape of pred.y: torch.Size([512, 1]) Shape of target: torch.Size([512, 0]) Can you please debug it? — Reply to this email directly, view it on GitHub <#68>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEJOOTW5T6YDNV2JAWZ72ZLX4OC47ANCNFSM6AAAAAA5ITIX3Y. You are receiving this because you are subscribed to this thread.
I
I met a similar problem for the guided generation. My solution is to denote the target as guidance_target in the file "regressor_model.yaml". I think Clement made a mistake here.
In fact the problem is that the config file "regressor_model.yaml" denote the target while the "qm9_dataset.py" denote it as cfg.general.guidance_target. The configuration did not perform well to extract the data features.
Hi, I met a question about the regressor model. I think the model does not correctly performed the conditional generation. The error for this command python3 src/guidance/train_qm9_regressor.py +experiment=regressor_model.yaml happens at the dimension of graph labels. Here is the error. target: tensor([], size=(512, 0)) Shape of pred.y: torch.Size([512, 1]) Shape of target: torch.Size([512, 0]) Can you please debug it?
Hi, I can't find the file:src/guidance/train_qm9_regressor.py
I met the same problem when I run "python3 src/guidance/train_qm9_regressor.py +experiment=regressor_model.yaml".
RuntimeError: Predictions and targets are expected to have the same shape, but got torch.Size([512, 1]) and torch.Size([512, 0]).