FusionNet_34Bx2_MoE
Fine-tuned model on English language using MoE method.
Model description
The FusionNet_34Bx2_MoE is a model to experiment with the MoE method, which could significantly increase the performance of the original model. The FusionNet_34Bx2_MoE has 60.8B parameters, and this model is fine-tuned. Enjoy!
Usage
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("TomGrc/FusionNet_34Bx2_MoE")
model = AutoModelForCausalLM.from_pretrained("TomGrc/FusionNet_34Bx2_MoE")
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 77.07 |
| AI2 Reasoning Challenge (25-Shot) | 72.95 |
| HellaSwag (10-Shot) | 86.22 |
| MMLU (5-Shot) | 77.05 |
| TruthfulQA (0-shot) | 71.31 |
| Winogrande (5-shot) | 83.98 |
| GSM8k (5-shot) | 70.89 |
- Downloads last month
- 987
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.
Model tree for TomGrc/FusionNet_34Bx2_MoE
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard72.950
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard86.220
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard77.050
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard71.310
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard83.980
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard70.890