| Quantization made by Richard Erkhov. | |
| [Github](https://github.com/RichardErkhov) | |
| [Discord](https://discord.gg/pvy7H8DZMG) | |
| [Request more models](https://github.com/RichardErkhov/quant_request) | |
| fairseq-dense-125M - bnb 8bits | |
| - Model creator: https://huggingface.co/KoboldAI/ | |
| - Original model: https://huggingface.co/KoboldAI/fairseq-dense-125M/ | |
| Original model description: | |
| --- | |
| language: en | |
| --- | |
| This is a Hugging Face transformers-compatible conversion of the original dense 125M-parameter model from the paper "[Efficient Large Scale Language Modeling with Mixtures of Experts](https://arxiv.org/abs/2112.10684)" from Artetxe et al. Please refer to the original model card, which can be found at https://github.com/facebookresearch/fairseq/blob/main/examples/moe_lm/model_card.md. | |
| # [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_KoboldAI__fairseq-dense-125M) | |
| | Metric | Value | | |
| |-----------------------|---------------------------| | |
| | Avg. | 26.0 | | |
| | ARC (25-shot) | 24.06 | | |
| | HellaSwag (10-shot) | 34.14 | | |
| | MMLU (5-shot) | 23.98 | | |
| | TruthfulQA (0-shot) | 43.72 | | |
| | Winogrande (5-shot) | 50.59 | | |
| | GSM8K (5-shot) | 0.0 | | |
| | DROP (3-shot) | 5.5 | | |