Model Stock: All we need is just a few fine-tuned models
Paper • 2403.19522 • Published • 14
merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the Model Stock merge method using bunnycore/Llama-3.1-8B-TitanFusion as a base.
Models Merged
The following models were included in the merge:
- DreadPoor/Heart_Stolen-8B-Model_Stock
- arcee-ai/Llama-3.1-SuperNova-Lite
- DreadPoor/Aspire-8B-model_stock
- vicgalle/Configurable-Llama-3.1-8B-Instruct
- mlabonne/Hermes-3-Llama-3.1-8B-lorablated
Configuration
The following YAML configuration was used to produce this model:
models:
- model: DreadPoor/Aspire-8B-model_stock
- model: arcee-ai/Llama-3.1-SuperNova-Lite
- model: mlabonne/Hermes-3-Llama-3.1-8B-lorablated
- model: bunnycore/Llama-3.1-8B-TitanFusion
- model: vicgalle/Configurable-Llama-3.1-8B-Instruct
- model: DreadPoor/Heart_Stolen-8B-Model_Stock
merge_method: model_stock
base_model: bunnycore/Llama-3.1-8B-TitanFusion
normalize: false
int8_mask: true
dtype: bfloat16
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 24.02 |
| IFEval (0-Shot) | 48.10 |
| BBH (3-Shot) | 32.07 |
| MATH Lvl 5 (4-Shot) | 12.99 |
| GPQA (0-shot) | 7.83 |
| MuSR (0-shot) | 11.94 |
| MMLU-PRO (5-shot) | 31.17 |
- Downloads last month
- 42
Model tree for bunnycore/Llama-3.1-8B-TitanFusion-v3
Merge model
this model
Paper for bunnycore/Llama-3.1-8B-TitanFusion-v3
Evaluation results
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard48.100
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard32.070
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard12.990
- acc_norm on GPQA (0-shot)Open LLM Leaderboard7.830
- acc_norm on MuSR (0-shot)Open LLM Leaderboard11.940
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard31.170