CoolFace
Modelpublic

trinhkhng/della_Merged_gpt2-medium_0.3

sourceHugging Faceupdated 2mo agoView on Hugging Face
0likes31downloads
Model Card

Mergedgpt2-medium0.3

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 /kaggle/working/gpt2-medium as a base.

Models Merged

The following models were included in the merge:

  • /kaggle/working/debias_gpt2-medium

Configuration

The following YAML configuration was used to produce this model:

yaml
base_model: /kaggle/working/gpt2-medium
dtype: float32
merge_method: della
models:
- model: /kaggle/working/debias_gpt2-medium
  parameters:
    density: 0.5
    epsilon: 0.1
    weight: 1.0
parameters:
  int8_mask: true
  lambda: 0.3
  normalize: true
  rescale: true
tokenizer:
  source: base