You will get Expert to train your custom LoRa model (qwen wan)
Project details
As an expert in AI model training, I provide professional-grade LoRA services designed for maximum likeness and flexibility. Unlike automated solutions, I leverage my expertise to manually curate your dataset with a strategic 70% close-up / 30% full-body split. I meticulously caption every image to ensure the AI understands exactly which features to keep consistent, delivering a high-fidelity model ready for any scenario
What's included
| Service Tiers |
Starter
$130
|
Standard
$230
|
Advanced
$330
|
|---|---|---|---|
| Delivery Time | 2 days | 2 days | 2 days |
Number of Revisions | 0 | 0 | 0 |
Model Validation/Testing | |||
Model Documentation | |||
Data Source Connectivity | |||
Source Code |
1 review
(1)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
LH
Lawrence H.
Feb 16, 2026
GenAI Photo and Video Creation for Dating Profile
About Yordan
Expert AI Creator
Plovdiv, Bulgaria - 7:46 pm local time
You don't need generic AI art. You need precise, controllable workflows that deliver consistent results. I specialize in the technical side of Generative AI—building complex ComfyUI workflows, training custom models, and implementing the newest papers/models the moment they are released.
My Technical Expertise:
🔹 Advanced ComfyUI Workflows:
Building custom node trees for automation and precision.
Mastery of Flux.1 (Pro/Dev/Schnell) and SDXL for high-end generation.
Implementation of Qwen-VL / Qwen Image for vision analysis and smart editing.
Z-Image / Zero-Shot workflows for instant character consistency without training.
🔹 Model Training & Fine-Tuning:
LoRA Training: Creating 99% accurate likeness models for Virtual Influencers.
Dataset curation, captioning, and regularization for perfect flexibility.
🔹 AI Video & Motion (SOTA):
Wan 2.1 / 2.x: Generating ultra-realistic AI video content.
Motion-controlled video using ControlNet and AnimateDiff.
🔹 Precision Editing:
Inpainting and Outpainting using Qwen Image Edit.
Face swapping and restoration (FaceID, IP-Adapter).
Why work with me?
I stay on the cutting edge. While others are still using Midjourney, I am building workflows with Wan 2.1 and Flux. I deliver clean, error-free .json workflows or final image/video assets ready for production.
Check my portfolio below to see consistent characters and advanced video examples.
Ready to build? Click Invite to Job.
Steps for completing your project
After purchasing the project, send requirements so Yordan can start the project.
Delivery time starts when Yordan receives requirements from you.
Yordan works on your project following the steps below.
Revisions may occur after the delivery date.
Training & Likeness Verification
I will train the initial LoRA model and generate a "grid" of test images. I will manually verify that the AI is capturing your facial identity correctly across different lighting conditions before finalizing the file.
Data Tagging & Model Training
I will manually caption your provided images, specifically tagging clothing and background elements to ensure they don't bleed into your identity. I then run the training process to create your custom file.