Browse categories
Explore
Fiverr Pro
English
$
USD
I will train a custom character or style lora
Kraken'it Media
About this gig
A LoRA is how you make an AI model that knows YOUR character same face, same outfit, every generation, forever. Most "AI artists" can't build one properly. I train them for a living.
**What makes mine different:**
Real dataset curation I hand-clean and re-caption every image so identity tags don't get overridden (the #1 reason amateur LoRAs drift)
Trained on a modern pipeline, with high-end hardware.
Multi-outfit support: your character in their signature outfit AND flexible for new ones
Delivered with trigger words, recommended weights, sampler settings, and a test grid proving consistency
Your files stay yours dataset and LoRA are never shared or reused
**Who buys this:** authors who want unlimited art of their book's hero, game devs, VTubers, brands with a mascot, and artists who want a LoRA of their own style (Standard/Premium with proof you own the artwork).
Not sure which package? Message me with what images you have and I'll tell you honestly what you need.
Delivery style preference
Please inform the freelancer of any preferences or concerns regarding the use of AI tools in the completion and/or delivery of your order.
Get to know Kraken'it Media
Kraken'it Media
Unleash the Kraken on your projects, fast, reliable, and done right!
- FromUnited States
- Member sinceMay 2024
- Avg. response time1 hour
Languages
English
Welcome to Kraken'it Media, Yes! I'm a real person(not just what an AI would say) where your tasks get tackled with precision, creativity, and tentacles of determination.
I specialize in fast-turnaround digital services including custom images/art, background removal, data entry, video subtitles, and basic editing. Whether you're a business needing clean visuals or a content creator wanting polished deliverables, I'm here to make it happen, on time and on point.
💡 Fast delivery
💬 Clear communication
⭐️ Detail-oriented results
Let’s get Kraken on your next task.
FAQ
What base models will it work with?
I can train on anything, just need to know what you're looking for.
I have zero images, just an idea.
That's the Premium package, I design the character with you first, then train.
Style LoRAs?
Yes, from YOUR OWN artwork only (30+ images recommended). I don't train on other artists' work without documented permission, no exceptions.
Real People?
Only yourself or someone who has given written consent (verification required). No celebrities, no exceptions.
What do I receive?
The .safetensors file, trigger word list, a settings/usage PDF, and the test grid. Standard+ includes the dataset.
NSFW capability?
I deliver the model plus SFW sample renders only. A LoRA is a neutral character model, how you use a model you own is your call, but I don't produce or deliver explicit content on Fiverr.
