FusionNet_7Bx2_MoE_v0.1
Fine-tuned model on English language using MoE method. The improved version from FusionNet_7Bx2_MoE_14B.
Model description
The FusionNet_7Bx2_MoE_v0.1 is a model to experiment with the MoE method, which could significantly increase the performance of the original model. The FusionNet has 12.9B parameters, and this model is fine-tuned. Enjoy!
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 76.16 |
| AI2 Reasoning Challenge (25-Shot) | 74.06 |
| HellaSwag (10-Shot) | 88.90 |
| MMLU (5-Shot) | 65.00 |
| TruthfulQA (0-shot) | 71.20 |
| Winogrande (5-shot) | 87.53 |
| GSM8k (5-shot) | 70.28 |
- Downloads last month
- 331
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.
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard74.060
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard88.900
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard65.000
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard71.200
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard87.530
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard70.280