🐮cow architecture gguf encoder

• don't need to rebuild tokenizer from metadata ⏳🥊
• don't need separate tokenizer file 🐱🥊
• no more oom issues (possibly) 💫💻🥊

eligible model example

• use cow-mistral3small 7.73GB for flux2-dev
• use cow-gemma2 2.33GB for lumina
• use cow-umt5base 451MB for ace-audio
• use cow-umt5xxl 3.67GB for wan-s2v or any wan video model