ml-training

Add training dataset from Supabase/local/HuggingFace

Add ML inference endpoint to FastAPI backend

---

Add ML UI components to Next.js frontend

Add parameter-efficient fine-tuning (LoRA/QLoRA/prefix-tuning)

Add cloud GPU platform integration (Modal/Lambda/RunPod)

Add data preprocessing pipelines (tokenization/transforms)

Create training configuration for classification/generation/fine-tuning

Deploy trained model for serverless inference

Deploy training job to cloud GPU platform

Estimate training and inference costs

Initialize ML training project with cloud GPU setup

Connect ML pipeline to Supabase storage

Monitor active training jobs and display metrics

Optimize training settings for cost and speed

Configure training framework (HuggingFace/PyTorch Lightning/Ray)

Test ML components (data/training/inference)

Validate training data quality and format