SinhalaMMLU-SharedTask/team-cognilex-qwen35-4b-synthetic-qa-it
Text Generation • 5B • Updated • 36
None defined yet.
Official submission organisation for the Sinhala MMLU Shared Task at ICTer 2026, Colombo, Sri Lanka.
Each team creates one repository in this organisation:
team-<teamname> (hyphens instead of spaces, e.g. team-sinhala-mmlu)
| File / folder | Contents |
|---|---|
inference.py (or .ipynb) |
The script that produced your final Codabench submission |
predictions.jsonl |
The exact file you submitted to the final leaderboard |
requirements.txt |
Packages and versions used |
README.md |
Team details, models, parameter count, GPU and run time |
training/ |
All training, fine-tuning and data-preparation code |
| data | Any extra data used for training, or links to it |
If you trained or fine-tuned a model, upload each model or adapter as its own repository: team-<teamname>-<model> (e.g. team-sinhala-mmlu-qwen-lora).
Register your team here to participate.
Questions: sinhalammlu.icter2026@gmail.com