Model Description
This is a model using the llama2 architecture and only 2 million parameters. It is trained on approximately 2 billion tokens of diverse web data from the first 1000000 rows of the uncleaned c4 english dataset.
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 3.46 |
| IFEval (0-Shot) | 15.82 |
| BBH (3-Shot) | 3.12 |
| MATH Lvl 5 (4-Shot) | 0.00 |
| GPQA (0-shot) | 0.00 |
| MuSR (0-shot) | 0.53 |
| MMLU-PRO (5-shot) | 1.26 |
- Downloads last month
- 18
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Dataset used to train cpayne1303/smallcp2024
Evaluation results
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard15.820
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard3.120
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard0.000
- acc_norm on GPQA (0-shot)Open LLM Leaderboard0.000
- acc_norm on MuSR (0-shot)Open LLM Leaderboard0.530
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard1.260