Phi
Collection
Microsoft-based models
•
107 items
•
Updated
•
1
merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the Passthrough merge method using Triangle104/Phi4-RP-o1 + ngxson/LoRA-phi-4-abliterated as a base.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
base_model: Triangle104/Phi4-RP-o1+ngxson/LoRA-phi-4-abliterated
dtype: bfloat16
merge_method: passthrough
models:
- model: Triangle104/Phi4-RP-o1+ngxson/LoRA-phi-4-abliterated
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 28.68 |
| IFEval (0-Shot) | 2.39 |
| BBH (3-Shot) | 51.22 |
| MATH Lvl 5 (4-Shot) | 38.82 |
| GPQA (0-shot) | 15.10 |
| MuSR (0-shot) | 18.93 |
| MMLU-PRO (5-shot) | 45.61 |
- Downloads last month
- 32
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for Triangle104/Phi4-RP-o1-Ablit
Collections including Triangle104/Phi4-RP-o1-Ablit
Evaluation results
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard2.390
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard51.220
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard38.820
- acc_norm on GPQA (0-shot)Open LLM Leaderboard15.100
- acc_norm on MuSR (0-shot)Open LLM Leaderboard18.930
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard45.610