merge
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the SLERP merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
slices:
- sources:
- model: 01-ai/Yi-1.5-34B-Chat
layer_range:
- 0
- 60
- model: BattlescarZa/medibuddy-llm-34B
layer_range:
- 0
- 60
merge_method: slerp
base_model: 01-ai/Yi-1.5-34B-Chat
parameters:
t:
- filter: self_attn
value:
- 0
- 0.5
- 0.3
- 0.7
- 1
- filter: mlp
value:
- 1
- 0.5
- 0.7
- 0.3
- 0
- value: 0.38
dtype: bfloat16
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 27.70 |
| IFEval (0-Shot) | 42.35 |
| BBH (3-Shot) | 42.81 |
| MATH Lvl 5 (4-Shot) | 12.24 |
| GPQA (0-shot) | 14.09 |
| MuSR (0-shot) | 15.97 |
| MMLU-PRO (5-shot) | 38.77 |
- Downloads last month
- 12
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 allknowingroger/Yibuddy-35B
Evaluation results
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard42.350
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard42.810
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard12.240
- acc_norm on GPQA (0-shot)Open LLM Leaderboard14.090
- acc_norm on MuSR (0-shot)Open LLM Leaderboard15.970
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard38.770