Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch
Paper
•
2311.03099
•
Published
•
30
This is a merge of pre-trained language models created using mergekit.
This model was merged using the DARE TIES merge method using appvoid/palmer-003 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: vihangd/DopeyTinyLlama-1.1B-v1
parameters:
density: 0.25
weight: 0.5
- model: raidhon/coven_tiny_1.1b_32k_orpo_alpha
parameters:
density: 0.75
weight: 0.5
merge_method: dare_ties
base_model: appvoid/palmer-003
parameters:
normalize: false
int8_mask: true
dtype: float16