Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time
Paper • 2203.05482 • Published • 10
# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("Ba2han/linear_merge")
model = AutoModelForCausalLM.from_pretrained("Ba2han/linear_merge", device_map="auto")Quick Links
p04_linear_90
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the Linear merge method using /ckpts/hf_repo/checkpoint-5500-v2 as a base.
Models Merged
The following models were included in the merge:
- /ckpts/hf_repo/checkpoint-18900
Configuration
The following YAML configuration was used to produce this model:
models:
- model: /ckpts/hf_repo/checkpoint-5500-v2
parameters:
weight: 0.09999999999999998
- model: /ckpts/hf_repo/checkpoint-18900
parameters:
weight: 0.9
merge_method: linear
base_model: /ckpts/hf_repo/checkpoint-5500-v2
parameters:
normalize: true
dtype: bfloat16
tokenizer_source: /ckpts/hf_repo/checkpoint-18900
- Downloads last month
- 7
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Ba2han/linear_merge")