Create README.md

Add num_batches_tracked buffers for HF transformers compatibility

Fix resample to BICUBIC (3) to match PaddleX preprocessing

Convert safetensors to HF format (support both paddle_dynamic and transformers backends)

Update safetensors with PaddleDet-format keys (no runtime conversion needed)

Upload preprocessor_config.json with huggingface_hub

Upload inference.yml with huggingface_hub

Upload config.json with huggingface_hub

Upload model.safetensors with huggingface_hub

Upload inference.yml with huggingface_hub

Upload config.json with huggingface_hub

initial commit