Fix Dockerfile: replace HHEM pre-warm with nli-deberta-v3-small

Fix Vectara label check and input format

Load T5-small tokenizer for Vectara HHEM v2

Use T5Tokenizer directly for Vectara HHEM v2

Fix Vectara pipeline: explicitly load tokenizer before pipeline init

Load Vectara model via transformers pipeline, not CrossEncoder

Speed up Docker builds: .dockerignore + merge model download layers

Add trust_remote_code=True for Vectara hallucination model

Switch faithfulness grader to Vectara hallucination evaluation model

Replace Anthropic with free-tier stack

Add full RAG evaluation pipeline with L1 metrics and UI