Migration: Switch to Gradio API (merve/fashn-vton-1.5) and updated token

Fix: Final alignment of dependencies and loading logic

Fix: Corrected StableVITONInference initialization and aligned logs

Fix: Force safetensors and fp16 variant to resolve missing text_encoder_2 weights

Fix: Pin transformers to 4.36.2 to resolve MT5Tokenizer import error

Deployment: Upgrade to modern dependencies for IDM-VTON support

Fix: Remove strict use_safetensors and handle variant fallback for model loading

Fix: Downgrade huggingface-hub to 0.25.2 for compatibility with diffusers 0.20.2

Fix: Use --extra-index-url to resolve PyPI dependencies

Deployment: Use official StableVITON requirements with pinned versions

Finalize requirements.txt with explicit sentencepiece for transformers compatibility

Fix: Update dependencies to resolve MT5Tokenizer import error

Redeploying backend with latest inference wrapper and test scripts

Trigger clean deployment v1.0.1

Use transformers[sentencepiece] and pin protobuf

Add sentencepiece and protobuf for MT5Tokenizer

Fix inference_wrapper.py imports and structure

Initial backend deployment