LLM Explainer v1.6.2

• Live batch progress header: done/running/queued counts, average per-function time, and a whole-batch-throughput ETA (reflects multiple llama-server endpoints running in parallel).
• Fix: BatchProgressDialog crashed on open because _update_progress_label was never defined.
• README: batch-mode screenshot with a local Qwen model on consumer hardware (Ryzen 9 9950X, 64 GB DDR5-6000, RX 7900 XTX 24 GB).