Is this copycat?

#1
by Jinsen322 - opened

You said

Each LoRA is paired with a single sample image of the same filename. You can drag the sample image into ComfyUI to load the workflow and parameters.

But this is lie, it doesnt work at all.

Also MIT license and can not be usef for commercial? Which one correct?

I know you tried to SEO hijack other creator, using same name and their hardwork. It's legal but shameless really.

I just hope your action dont bring any bad outcome in future.

I don't remember seeing any of these names in the other repo so I think the only thing being copied here is the name of the repo, which is rather generic if we're being honest. Also the loras do work even if the workflows aren't embedded in the sample images. Just use them in whatever generic Krea 2 workflow you use.

These LoRAs were trained on my local computer, and honestly, it's my very first time training loras. The quality still has room for improvement. Not a copycat.

But why use same title and even copied the readme? What's your idea behind of it?

People will get confused

Hi sorry my friend jinsen322 is being stingy here, you can keep the repo name

I will just change mine and you can keep yours, if you want to get any help on training lora just tell me.

What I recommend is to use around 25-30 images with good quality, also train around 2000-2500 step go get epoch around 25-33.

The key is here epoch, use clean image too, artifact will ruin the model

@Jinsen322 Regarding the name, it's more like this: Me: 'Hey XXX[bot], pls study the popular repo names and READMEs on HF, and write me one based on what you learned and what our LoRAs are....' XXX[bot]: 'No Problem! There you go...' And to be honest, it's quite catchy and understandable.

Hey @fidel1234 thx. Just to avoid any misunderstanding, I'll come up with a new name later and avoid being anything like yours, thx again for your advice!

The original creator said to use it, he already change it so keep it

Use this one


job: "extension"
config:
name: "krea_template"
process:
- type: "diffusion_trainer"
training_folder: ""
sqlite_db_path: "./aitk_db.db"
device: "cuda"
trigger_word: null
performance_log_every: 10
network:
type: "lora"
linear: 32
linear_alpha: 32
lokr_full_rank: true
lokr_factor: -1
network_kwargs:
ignore_if_contains: []
save:
dtype: "bf16"
save_every: 250
max_step_saves_to_keep: 4
save_format: "diffusers"
push_to_hub: false
datasets:
- folder_path: ""
mask_path: null
mask_min_value: 0.1
default_caption: ""
caption_ext: "txt"
caption_dropout_rate: 0.05
cache_latents_to_disk: false
is_reg: false
network_weight: 1
resolution:
- 512
- 768
- 1024
controls: []
shrink_video_to_frames: true
num_frames: 1
flip_x: false
flip_y: false
num_repeats: 1
train:
batch_size: 1
bypass_guidance_embedding: false
steps: 2500
gradient_accumulation: 1
train_unet: true
train_text_encoder: false
gradient_checkpointing: true
noise_scheduler: "flowmatch"
optimizer: "adamw8bit"
timestep_type: "sigmoid"
content_or_style: "balanced"
optimizer_params:
weight_decay: 0.0001
unload_text_encoder: false
cache_text_embeddings: true
lr: 0.0001
ema_config:
use_ema: false
ema_decay: 0.99
skip_first_sample: true
force_first_sample: false
disable_sampling: true
dtype: "bf16"
diff_output_preservation: false
diff_output_preservation_multiplier: 1
diff_output_preservation_class: "person"
switch_boundary_every: 1
loss_type: "mse"
logging:
log_every: 1
use_ui_logger: true
model:
name_or_path: "krea/Krea-2-Raw"
quantize: true
qtype: "qfloat8"
quantize_te: true
qtype_te: "qfloat8"
arch: "krea2"
low_vram: false
model_kwargs: {}
compile: false
layer_offloading: false
layer_offloading_text_encoder_percent: 1
layer_offloading_transformer_percent: 1
sample:
sampler: "flowmatch"
sample_every: 99999
sample_start_step: 0
width: 1024
height: 1024
samples:
- prompt: "woman with red hair, playing chess at the park, bomb going off in the background"
- prompt: "a woman holding a coffee cup, in a beanie, sitting at a cafe"
- prompt: "a horse is a DJ at a night club, fish eye lens, smoke machine, lazer lights, holding a martini"
- prompt: "a man showing off his cool new t shirt at the beach, a shark is jumping out of the water in the background"
- prompt: "a bear building a log cabin in the snow covered mountains"
- prompt: "woman playing the guitar, on stage, singing a song, laser lights, punk rocker"
- prompt: "hipster man with a beard, building a chair, in a wood shop"
- prompt: "photo of a man, white background, medium shot, modeling clothing, studio lighting, white backdrop"
- prompt: "a man holding a sign that says, 'this is a sign'"
- prompt: "a bulldog, in a post apocalyptic world, with a shotgun, in a leather jacket, in a desert, with a motorcycle"
neg: ""
seed: 42
walk_seed: true
guidance_scale: 4
sample_steps: 30
num_frames: 1
fps: 1
meta:
name: "[name]"
version: "1.0"

Basicly ai toolkit krea 2 default and use 2500 step (around 25-30 images)
Sigmoid + cache text embedded, you can try automagic v3 too

if you ad more images = more training step
fewer images = lesser training step

Avoid to use same similar picture

@fidel1234 thx again for your tips, quite helpful actually.

All misunderstandings are cleared up, I'm gonna close the comment now.

lorayap changed discussion status to closed

Sign up or log in to comment