multimodalart/krea2-lora-trainer
Set login button label: Sign in with Hugging Face to use the UI
Reword agent hint: paste this to your coding agent
Remove dead submit_from_dataset route helper
Revert to demo.launch() + file-based agent recipe (drop FastAPI/custom routes)
Add standalone train_job.py for agents
Add agent recipe (agent.md) for curl-based LoRA training
agent: standalone CLI job script + submit_from_dataset; Krea token now optional
agent: host /agent recipe + /train_job.py + /api/train; mount via FastAPI
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload app.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload app.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload app.py with huggingface_hub
Upload jobs.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload caption.py with huggingface_hub
Upload app.py with huggingface_hub
Upload README.md with huggingface_hub
Upload jobs.py with huggingface_hub
Upload app.py with huggingface_hub
Upload app.py with huggingface_hub
Upload README.md with huggingface_hub
Upload requirements.txt with huggingface_hub
Upload caption.py with huggingface_hub
Upload app.py with huggingface_hub
Add optional Turbo validation toggle
Fix UV git dependency format (pkg @ git+url)
Krea 2 LoRA trainer (HF Jobs backend)
initial commit