Skip to content

feat: Add Andreas' first test submission #419

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
AndreasBis wants to merge 1 commit into
base: main
Choose a base branch
from
+2,682 −0
Open

feat: Add Andreas' first test submission #419

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
6 changes: 6 additions & 0 deletions records/track_non_record_16mb/first-submission/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
# Andreas' First Submission (Test)

This is a test submission in the non-record track. It uses the current SOTA architecture (10L Int5-MLP + BigramHash + SWA) to verify the submission workflow and gain experience.

### Setup
Ensure you install the dependencies listed in `requirements.txt`.
11 changes: 11 additions & 0 deletions records/track_non_record_16mb/first-submission/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,11 @@
numpy
tqdm
torch
huggingface-hub
kernels
setuptools
typing-extensions==4.15.0
datasets
tiktoken
sentencepiece
zstandard
9 changes: 9 additions & 0 deletions records/track_non_record_16mb/first-submission/submission.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
{
"name": "Andreas Test Submission",
"val_loss": 1.14276,
"bytes_total": 15900000,
"blurb": "Test submission based on the first place solution (10L Int5-MLP).",
"author": "Andreas",
"github_id": "andreasbis",
"date": "2026-03-22"
}
Comment on lines +1 to +9
Copy link

Copilot AI Mar 22, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This submission.json format doesn’t match the other entries in records/track_non_record_16mb, which include track and typically record both val_loss and val_bpb. Also, val_loss here is set to ~1.14, but the accompanying train.log shows val_loss ~1.929 and val_bpb ~1.143 after quant roundtrip, so the metric appears mislabeled. Recommend adding track (e.g. non-record-16mb), adding val_bpb, and setting val_loss/bytes_total to the exact values emitted by the training log.

Copilot uses AI. Check for mistakes.
Loading
Loading