Re-export with coreai-torch 0.4.1 (fixes OS 27 beta 2+ load failure); gated token-for-token vs fp32 oracle

Note b2 (beta3-loadable) bundles under gpu-pipelined-b2/

Add b2 (beta3-loadable) bundle under gpu-pipelined-b2/youtu_llm_2b_decode_absorbed_int8_msdpa_g32/ (b1 retained)

Card: lead with CoreAIKit ChatSession (kit-enrolled)

Add source config.json

Add model card

Add Youtu-LLM-2B int8 pipelined-decode bundle (absorbed MLA)

initial commit