license: apache-2.0 | |
This is just a standard conversion to hf transformers format of models from here: | |
https://huggingface.co/BlinkDL/rwkv-4-pileplus | |
According to the documentation I found, this model should have seen roundabout 2 trillion tokens! | |
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) | |
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_KnutJaegersberg__RWKV-4-PilePlus-169M-20230520-done-ctx4096) | |
| Metric | Value | | |
|-----------------------|---------------------------| | |
| Avg. | 25.0 | | |
| ARC (25-shot) | 23.98 | | |
| HellaSwag (10-shot) | 32.25 | | |
| MMLU (5-shot) | 23.37 | | |
| TruthfulQA (0-shot) | 42.29 | | |
| Winogrande (5-shot) | 49.17 | | |
| GSM8K (5-shot) | 0.38 | | |
| DROP (3-shot) | 3.53 | | |