Model Stock: All we need is just a few fine-tuned models
Paper
• 2403.19522 • Published
• 13
This is a merge of pre-trained language models created using mergekit.
This model was merged using the Model Stock merge method using sthenno-com/miscii-14b-0218 as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
name: tempesthenno-ms-0314-001
merge_method: model_stock
base_model: sthenno-com/miscii-14b-0218
tokenizer:
source: base
dtype: bfloat16
parameters:
normalize: true
rescale: false
models:
- model: sthenno/tempesthenno-sft-0309-ckpt10
- model: sthenno/tempestissimo-14b-0309
- model: /home/ubuntu/tmp/models/tempesthenno-sft-0314-stage1-ckpt50
- model: /home/ubuntu/tmp/models/tempesthenno-sft-0314-stage1-ckpt100
- model: /home/ubuntu/tmp/models/tempesthenno-sft-0314-stage3-ckpt30