Fix benchmark section: remove placeholder 0.0/10 scores

Update README: add collection link

Add GitHub research project link; remove unverified benchmark scores

Add Nondual Bot Benchmark v1.0 eval results (50q, 10 categories)

Upload Atmananda_Agent-Q8_0.gguf with huggingface_hub

Add disclaimer section: independent research, weights-only, no affiliation

Update model card with retrain results (full corpus, no opener-cap)

Remove undertrained GGUF (opener-cap bug, retrain in progress)

Add model card with training details, examples, and methodology

Upload Atmananda_Agent-Q8_0.gguf with huggingface_hub

initial commit