bugkira-ai/babylm-paraslstm-20m

Card cleanup: sister links, drop Apple mention / fluff

Link sister ParaNLRU BabyLM model

Fix tokenizer.json: ByteLevel decoder for readable decode

fix config.json: generation without KV cache

fix configuration_babylm_pararnn.py: generation without KV cache

fix modeling_babylm_pararnn.py: generation without KV cache

fix README.md: self-contained modeling (pararnn only) + generate

fix modeling_babylm_pararnn.py: self-contained modeling (pararnn only) + generate

Add ParaSLSTM BabyLM-10M Strict-Small PoC (shuf ep3) + model card

initial commit