Upload Llama-3.2-3B-Instruct-cuda-q8.base with huggingface_hub

Upload Llama-3.2-3B-Instruct-cuda-q4mix.base with huggingface_hub

Re-convert with fixed rotary q/k permutation (baseRT-internal#114); NIAH-validated at 2k/8k. Q4 now uses the default-q4 quality profile (embeddings/head at higher precision).

Re-convert with fixed rotary q/k permutation (baseRT-internal#114); NIAH-validated at 2k/8k. Q4 now uses the default-q4 quality profile (embeddings/head at higher precision).

Add config.json (enables HF download tracking)

Upload README.md with huggingface_hub

Upload Llama-3.2-3B-Instruct-Q8.base with huggingface_hub

Upload Llama-3.2-3B-Instruct-Q4.base with huggingface_hub

initial commit