djuna/Q2.5-Partron-7B

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the della merge method using djuna/Q2.5-Fuppavy-7B as a base.

Models Merged

The following models were included in the merge:

• Locutusque/StockQwen-2.5-7B
• fblgit/cybertron-v4-qw7B-MGS
• happzy2633/qwen2.5-7b-ins-v3

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: Locutusque/StockQwen-2.5-7B
    parameters:
      weight: 0.5
      density: 0.5
  - model: happzy2633/qwen2.5-7b-ins-v3
    parameters:
      weight: 0.3
      density: 1
  - model: fblgit/cybertron-v4-qw7B-MGS
    parameters:
      weight: 1
      density: 0.8
merge_method: della
base_model: djuna/Q2.5-Fuppavy-7B
parameters:
  epsilon: 0.04
  lambda: 1.05
dtype: float32
out_dtype: bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	27.08
IFEval (0-Shot)	73.21
BBH (3-Shot)	35.26
MATH Lvl 5 (4-Shot)	0.08
GPQA (0-shot)	6.38
MuSR (0-shot)	11.07
MMLU-PRO (5-shot)	36.47