fix(config): sync with upstream Google Gemma 4 day-1 fixes (b446025, b4a6011)

tokenizer_config: fix response schema regex — tool_calls before content, add tool_response terminator (from google upstream 2026-04-10)

docs: correct base_model lineage for HF model tree

feat: add gguf multi-quant set (migrated from lthn/lemmy-gguf)

cleanup: drop mlx files, align with lemer gguf-playground pattern

eval: first 8-PAC paired run — strong LEK delta signal

docs: update README for LEK-merged Lemma family

feat: merge LEK into lemmy weights

fix: pull Google's chat_template.jinja fix (26B/31B variant)

fix: replace LICENSE symlink with real EUPL-1.2 file

Upload config.json with huggingface_hub

Delete files lemmy-bf16.gguf lemmy-q4_k_m.gguf lemmy-q5_k_m.gguf lemmy-q6_k.gguf lemmy-q8_0.gguf with huggingface_hub

Upload tokenizer_config.json with huggingface_hub

Upload config.json with huggingface_hub

Update README.md

Update README.md

Upload processor_config.json with huggingface_hub

Upload config.json with huggingface_hub

Update README.md

Upload folder using huggingface_hub

Replace bf16 with Q4 MLX — consumer-ready quantisation

Upload Lemmy 26B bf16 (MLX format) — LEK-activated Gemma 4 26B-A4B MoE

initial commit