Skip to content

fix PriorDumpDataLoader to correctly handle num_datapoints #9

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

fix PriorDumpDataLoader to correctly handle num_datapoints #9

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e88425b
fix PriorDumpDataLoader to correctly handle num_datapoints and max_se…
kursatfelsen Oct 6, 2025
d077f0d
Update nanotabpfn/priors.py
kursatfelsen Oct 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions nanotabpfn/priors.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,9 +28,10 @@ def __init__(self, filename, num_steps, batch_size, device, starting_index=0):
self.num_steps = num_steps
self.batch_size = batch_size
with h5py.File(self.filename, "r") as f:
self.num_datapoints_max = f['X'].shape[0]
self.num_datapoints_max = f['X'].shape[1]
self.max_num_classes = f['max_num_classes'][0]
self.problem_type = f['problem_type'][()].decode('utf-8')
self.has_num_datapoints = "num_datapoints" in f
self.device = device
self.pointer = starting_index

Expand All @@ -41,8 +42,12 @@ def __iter__(self):
end = self.pointer + self.batch_size

num_features=self.data['num_features'][self.pointer:end].max()
x = torch.from_numpy(self.data['X'][self.pointer:end,:,:num_features])
y = torch.from_numpy(self.data['y'][self.pointer:end])
if self.has_num_datapoints:
max_seq_in_batch = int(self.data['num_datapoints'][self.pointer:end].max())
else:
max_seq_in_batch = int(self.num_datapoints_max)
x = torch.from_numpy(self.data['X'][self.pointer:end, :max_seq_in_batch, :num_features])
y = torch.from_numpy(self.data['y'][self.pointer:end, :max_seq_in_batch])
single_eval_pos = self.data['single_eval_pos'][self.pointer:end]

self.pointer += self.batch_size
Expand Down