Merge this weights with original model or how can I prompt with the Fine Tuned weight?

#31
by Tahmimhossain - opened

Hi I have already Fine-Tuned the model and I got the "adapter_config.json" & "adapter_model.safetensors" files. But how I can merge this weights with original model or how can I prompt with the Fine Tuned weight?

https://huggingface.co/docs/peft/v0.12.0/en/developer_guides/checkpoint#merge-the-weights

Generic example

config = PeftConfig.from_pretrained("stevhliu/vit-base-patch16-224-in21k-lora")
model = AutoModelForImageClassification.from_pretrained(
    config.base_model_name_or_path,
    label2id=label2id,
    id2label=id2label,
    ignore_mismatched_sizes=True,
)
model = PeftModel.from_pretrained(model, "stevhliu/vit-base-patch16-224-in21k-lora")

merged_model = model.merge_and_unload()
This comment has been hidden
Tahmimhossain changed discussion status to closed

https://huggingface.co/docs/peft/v0.12.0/en/developer_guides/checkpoint#merge-the-weights

Generic example

config = PeftConfig.from_pretrained("stevhliu/vit-base-patch16-224-in21k-lora")
model = AutoModelForImageClassification.from_pretrained(
    config.base_model_name_or_path,
    label2id=label2id,
    id2label=id2label,
    ignore_mismatched_sizes=True,
)
model = PeftModel.from_pretrained(model, "stevhliu/vit-base-patch16-224-in21k-lora")

merged_model = model.merge_and_unload()

Thank this worked.

Sign up or log in to comment