← ImageLoRA
How to train a Fibo 2.0 LoRA
Which trainers support Fibo 2.0 / Edit Turbo (Bria, 8B, CC BY-NC 4.0), how much VRAM you need, and settings that work. Every row links to its source.
Status: Released Oct 20, 2026. BriaFibo2Pipeline is in diffusers main (no release yet, so install diffusers from source). Trainer support is pending: ai-toolkit, OneTrainer, musubi-tuner, kohya sd-scripts and SimpleTuner have no Fibo 2.0 support yet (checked Oct 21). You can train LoRAs today with Bria's own trainer, the diffusers DreamBooth scripts (open PR) or AcademiaSD LoRAlab. Non-commercial license: commercial use needs a paid agreement with Bria.
Fibo 2.0 LoRA: what to input (Bria's starting point)
| Source | Bria's fine_tuning guide (huggingface.co/briaai/Fibo-2.0-dev), summarized in diffusers PR #15084 |
|---|
| Trainer | Bria fine_tuning/train_lora.py (single GPU, text-to-image and edit), or diffusers train_dreambooth_lora_bria_fibo2.py (PR #15084) |
|---|
| Base model | Fibo 2.0 Dev (not Edit Turbo) |
|---|
| Network | LoRA, rank 64, alpha 64, all linear layers |
|---|
| Learning rate | 1e-4, constant |
|---|
| Resolution | 1024 (~1 MP) |
|---|
| Batch | 1, gradient accumulation 8 |
|---|
| Steps | 1,000 (diffusers t2i example) / 2,000 (edit example); likeness shows by about step 300 |
|---|
| Captions | Structured JSON (VGL, recommended in Bria's README) or plain sentences; both give the same result in practice |
|---|
| Save VRAM | cache latents, gradient checkpointing, 8-bit Adam, offload VAE and text encoder |
|---|
Trainer support
| Trainer | Status |
|---|
| ai-toolkit | unknown (no Fibo 2.0 support or PR yet, Oct 21) |
| OneTrainer | unknown (no Fibo 2.0 support or PR yet, Oct 21) |
| musubi-tuner | unknown (no Fibo 2.0 support or PR yet, Oct 21) |
| kohya sd-scripts | unknown (no Fibo 2.0 support or PR yet, Oct 21) |
| SimpleTuner | unknown (no Fibo 2.0 support or PR yet, Oct 21) |
VRAM
LoRA training: ~8.7 GB at 512px, rank 8 on an RTX 5080 with AcademiaSD LoRAlab's NF4 build. Weight sizes: transformer 16.45 GB bf16 or 8.41 GB int8 ConvRot; Qwen3-VL-4B text encoder 8.88 GB bf16 or 4.86 GB int8; FLUX.2 VAE 0.34 GB. Bria's self-hosted inference package asks for a 40 GB GPU. Forge edit plugin starting points: 12 GB at 512-768px, 16 GB at 768-1024px (estimated on an RTX 3090, not measured on 12/16 GB cards). Full bf16 LoRA training at 1024px: unknown
Known-good settings
- License: CC BY-NC 4.0, non-commercial use only. Commercial use (client work, paid gigs, products) needs a separate paid agreement with Bria; LoRAs trained on it are derivatives under the same terms
- Gated on Hugging Face: fill in Bria's access form (company, intended use) on the model page or downloads fail
- Train on Fibo 2.0 Dev. Edit Turbo is distilled for 4-step editing and is not meant for fine-tuning
- One set of weights does text-to-image and editing, so a single LoRA can learn a style or domain and how to edit within it
- Bria's own trainer: fine_tuning/train_lora.py in the Fibo-2.0-dev repo is a single-GPU LoRA trainer for both text-to-image and edit LoRAs; its README recommends JSON captions
- Bria's suggested LoRA start: rank 64 (alpha = rank), LR 1e-4 constant, 1024px, gradient accumulation 8, LoRA on every linear layer
- Captions: Fibo 2.0 is trained on structured JSON (VGL) captions; caption your dataset in the form you will prompt with. Plain sentences work and are used as the short_description. bria-vgl writes the JSON from a short prompt (it calls Gemini, so it needs a GEMINI_API_KEY)
- In practice JSON captions and plain captions give the same result, so plain captions are fine for a first run
- Training on Dev shows likeness by about step 300. Realistic subjects train well; anime results are mixed
- Edit LoRAs train on (source image, target image, instruction) triplets; the diffusers edit trainer raises max_sequence_length to 1024 because the source image's vision tokens count toward it (~576 tokens at 768px)
- diffusers LoRA trainers (PR #15084, still open, waiting on a rebase): train_dreambooth_lora_bria_fibo2.py (text-to-image) and _img2img.py (edit); transformer-only LoRA, text encoder and VAE frozen; needs transformers>=5.20.1. Memory flags: --cache_latents, --gradient_checkpointing, --use_8bit_adam, --offload, bitsandbytes quantized training, --do_fp8_training
- diffusers: BriaFibo2Pipeline is in main since Oct 20 (PR #14991) but not in a release yet (latest v0.41.0), so install diffusers from source (pip install git+https://github.com/huggingface/diffusers)
- AcademiaSD LoRAlab Trainer Studio trains Fibo 2.0 LoRAs from 8 GB using an NF4 build of Dev, and exports LoRAs in ComfyUI's format
- Resolution: ~1 MP, sides in multiples of 16 (1024x1024, 1424x800 for 16:9, 1248x832 for 3:2)
- Inference, Dev: 50 steps, CFG 5 (diffusers card); ComfyUI's native support uses Dev 30 steps, CFG 5 and Edit Turbo 4 steps, CFG 1, euler/simple
- Dev in ComfyUI that works: 50 steps, CFG 5, euler/simple, no negative prompt, 1 MP auto-resize
- Edits take a mask (white = regenerate) and up to 4 reference images, referred to as <image_1>, <image_2>...
- Edit mode is weak at small edits such as removing makeup and tends to redraw the whole head; describe the hair in the prompt to keep it
- Two-reference prompts work well: "the woman in image 1 and image 2 is now ..."
- ComfyUI native support (Dev and Edit Turbo) is on master (PR #17240, merged Oct 21), not in a stable release yet, so use the latest nightly/beta. Repackaged files: Comfy-Org/Fibo-2.0. ComfyUI-Fibo2 custom nodes run Edit Turbo on ComfyUI 0.42.0 (tested on an RTX 4090)
- ComfyUI: the int8 ConvRot Qwen3-VL text encoder errors with vision inputs (issue #17248, no fix yet); use the bf16 or fp8 scaled text encoder
- Architecture: 8B flow-matching single-stream DiT (44 blocks), FLUX.2 VAE, Qwen3-VL-4B-Instruct text encoder with a Perceiver that compresses it to 256-512 gist tokens
- Artificial Analysis (Oct 20): #1 open-weights model for image editing (Elo 1087, just ahead of Qwen-Image-2.1 at 1083) and #2 open-weights for text-to-image behind Qwen-Image-2.1
- Trained only on licensed data (~1B-image Bria dataset)
- Community take: Edit Turbo is very fast at 4 steps and good with references and object removal, but struggles with complex prompts and several references at once, and outputs stay around 1 MP. Many people prefer MOSS Image or Krea 2 for text-to-image looks
General tips
- Gated Hugging Face models: click 'Agree and access' on each model page, or you'll get 403 errors
- Steps scale with dataset size (steps = epochs x images at batch 1)
- A safe start: LoKr with auto rank (~64), batch/grad accumulation 4
Where to rent a GPU
See cheapest GPU to train a LoRA (RunPod, Vast.ai).
Sources
Weekly newsletter: new models, trainer releases, cheapest GPUs.
Subscribe
Data updated 2026-10-20. RunPod and Vast.ai links are referral links. unknown = no data yet. · Raw data (JSON) · llms.txt