This is a merge of pre-trained language models created using
mergekit.
This model was merged using the linear
DARE merge method using
mllm-dev/gpt2_f_experiment_0 as a base.
1base_model:
2 model:
3 path: mllm-dev/gpt2_f_experiment_0
4dtype: float16
5merge_method: dare_linear
6parameters:
7 normalize: 1.0
8slices:
9- sources:
10 - layer_range: [0, 12]
11 model:
12 model:
13 path: mllm-dev/merge_diff_data_DROID
14 parameters:
15 weight: 0.5
16 - layer_range: [0, 12]
17 model:
18 model:
19 path: mllm-dev/merge_diff_data_YELP
20 parameters:
21 weight: 0.5
22 - layer_range: [0, 12]
23 model:
24 model:
25 path: mllm-dev/gpt2_f_experiment_0
