FusionNet_34Bx2_MoE_v0.1
Fine-tuned model on English language using MoE method. The improved version from FusionNet_34Bx2_MoE.
Model description
The FusionNet_34Bx2_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 60.8B parameters, and this model is fine-tuned. Enjoy!
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 77.38 |
| AI2 Reasoning Challenge (25-Shot) | 73.72 |
| HellaSwag (10-Shot) | 86.46 |
| MMLU (5-Shot) | 76.72 |
| TruthfulQA (0-shot) | 71.01 |
| Winogrande (5-shot) | 83.35 |
| GSM8k (5-shot) | 73.01 |
- Downloads last month
- 63
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for TomGrc/FusionNet_34Bx2_MoE_v0.1
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard73.720
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard86.460
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard76.720
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard71.010
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard83.350
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard73.010