Modelpublic

je-suis-tm/abigail_spencer_lora_flux_nf4

sourceHugging Facemitupdated 9mo agoView on Hugging Face

0likes10downloads

Model Card

Abigail Spencer Lora Flux NF4

All files are also archived in https://github.com/je-suis-tm/huggingface-archive in case this gets censored.

The QLoRA fine-tuning process of abigail_spencer_lora_flux_nf4 takes inspiration from this post (https://huggingface.co/blog/flux-qlora). The training was executed on a local computer with 1000 timesteps and the same parameters as the link mentioned above, which took around 6 hours on 8GB VRAM 4060. The peak VRAM usage was around 7.7GB. To avoid running low on VRAM, both transformers and text_encoder were quantized. All the images generated here are using the below parameters

—Height: 512
—Width: 512
—Guidance scale: 5
—Num inference steps: 20
—Max sequence length: 512
—Seed: 0

Usage

python

import torch
from diffusers import FluxPipeline, FluxTransformer2DModel
from transformers import T5EncoderModel

text_encoder_4bit = T5EncoderModel.from_pretrained(
    "hf-internal-testing/flux.1-dev-nf4-pkg", subfolder="text_encoder_2",torch_dtype=torch.float16,)

transformer_4bit = FluxTransformer2DModel.from_pretrained(
        "hf-internal-testing/flux.1-dev-nf4-pkg", subfolder="transformer",torch_dtype=torch.float16,)

pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.float16,
                                    transformer=transformer_4bit,text_encoder_2=text_encoder_4bit)

pipe.load_lora_weights("je-suis-tm/abigail_spencer_lora_flux_nf4",
                       weight_name='pytorch_lora_weights.safetensors')

prompt="Full body view of Abigail Spencer dressed up as commanding Native Caribbean pirate captain. She is stands tall amidst the Pirate City of Lost Ships, her fierce expression and intricate tattoos reflecting chaos. Dramatic side lighting adds intensity to the scene, blending natural and futuristic elements. Wearing traditional pirate Captains Clothing and hat as depicted as Captain Jack Sparrow"

image = pipe(
            prompt,
            height=512,
            width=512,
            guidance_scale=5,
            num_inference_steps=20,
            max_sequence_length=512,
            generator=torch.Generator("cpu").manual_seed(0),            
        ).images[0]

image.save("abigail_spencer_lora_flux_nf4.png")

Trigger words

You should use Abigail Spencer to trigger the image generation.

Download model

Download them in the Files & versions tab.