Skip to content

Change TrivialMemoryModel to ResidualMemoryModel #2

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
PierreQuinton wants to merge 3 commits into
base: add-rnn-training
Choose a base branch
from
Open

Change TrivialMemoryModel to ResidualMemoryModel #2

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
ba590dc
Change TrivialMemoryModel to ResidualMemoryModel
PierreQuinton Dec 19, 2025
f454a66
Interesting, this seems to explode even with an insanely low LR.
PierreQuinton Dec 19, 2025
48f246c
Witb k=3 it works sometimes.
PierreQuinton Dec 20, 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
Next Next commit
Change TrivialMemoryModel to ResidualMemoryModel
  • Loading branch information
@PierreQuinton
PierreQuinton committed Dec 19, 2025
commit ba590dca2d19258fa9b5b07aeb3cd5706c668de8
6 changes: 3 additions & 3 deletions src/recursion/models/trivial_memory_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@
from recursion.dataset.repeat_after_k import make_sequences


class TrivialMemoryModel(nn.Module):
class ResidualMemoryModel(nn.Module):
def __init__(self, memory_dim: int):
super().__init__()

Expand All @@ -32,7 +32,7 @@ def forward(self, input: Tensor, memory: Tensor) -> tuple[Tensor, Tensor]:
x = self.relu(self.fc1(x))
x = self.fc2(x)

return x
return x + memory


batch_size = 16
Expand All @@ -41,7 +41,7 @@ def forward(self, input: Tensor, memory: Tensor) -> tuple[Tensor, Tensor]:

memory_dim = 8

model = TrivialMemoryModel(memory_dim)
model = ResidualMemoryModel(memory_dim)
head = nn.Linear(memory_dim, 1)
memory = torch.zeros(batch_size, memory_dim)
criterion = nn.BCEWithLogitsLoss(reduction="none")
Expand Down