IceMochaccinoRP-7b

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

Merge Details

For mergers!

Merge Method

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

• IceCappuccinoRP-multi_verse_model
• IceCappuccinoRP-IceCoffeeRP-7b

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - model: IceCappuccinoRP-IceCoffeeRP-7b
        layer_range: [0, 32]
      - model: IceCappuccinoRP-multi_verse_model
        layer_range: [0, 32]

merge_method: slerp
base_model: IceCappuccinoRP-multi_verse_model
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: 0.5 # fallback for rest of tensors
dtype: bfloat16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	68.95
AI2 Reasoning Challenge (25-Shot)	68.00
HellaSwag (10-Shot)	85.41
MMLU (5-Shot)	62.78
TruthfulQA (0-shot)	56.22
Winogrande (5-shot)	80.58
GSM8k (5-shot)	60.73