ComfyUI Node: Rebase LoRA From Finetune

Authored by srv1n

Created

Updated

0 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Category

model_merging/lora

Inputs

base_model MODEL
finetuned_model MODEL
base_lora_name
    save_relpath STRING
    rank INT
    layer_scope
    • original_lora_only
    • all_compatible_weights
    rank_mode
    • match_input_lora_when_possible
    • fixed
    output_mode
    • standard
    • full_diff_patch
    strength_model FLOAT
    strength_clip FLOAT
    save_dtype
    • fp16
    • bf16
    • fp32
    bias_diff BOOLEAN
    clamp_quantile FLOAT
    include_clip BOOLEAN
    compute_device
    • auto
    • cpu
    • cuda
    svd_mode
    • auto
    • full
    • lowrank
    cuda_svd_driver
    • auto
    • gesvdj
    • gesvda
    • gesvd
    lowrank_oversample INT
    lowrank_niter INT
    base_clip CLIP
    finetuned_clip CLIP

    Outputs

    STRING

    STRING

    Extension: ComfyUI-Rebase-LoRA

    This custom node builds a new LoRA or patch file relative to a finetuned model, computing weight differences and writing results as standard LoRA or exact diff patch files.

    Authored by srv1n

    Looking for a different node?

    Run ComfyUI workflows without the setup

    No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

    Learn more