Update README.md

README.md

@@ -27,16 +27,13 @@ library_name: transformers
   </div>
 <h1 style="margin-top: 0rem;">🌙 Kimi K2 Usage Guidelines</h1>
 </div>
-It is recommended to have at least 128GB unified RAM memory to run the small quants. Expect 5+ tokens/sec.
+It is recommended to have at least 128GB unified RAM memory to run the small quants. With 16GB VRAM and 256 RAM, expect 5+ tokens/sec.
 For best results, use any 2-bit XL quant or above.
 
 Set the temperature to 0.6 recommended) to reduce repetition and incoherence.
 
-For llama.cpp / GGUF inference, you should skip the BOS since it’ll auto add it:
-```
-<｜User｜>What is 1+1?<｜Assistant｜>
-```
-- For complete detailed instructions, see our guide: [unsloth.ai/blog/kimi-k2](https://docs.unsloth.ai/basics/kimi-k2)
+- Use llama.cpp's [PR #14654](https://github.com/ggml-org/llama.cpp/pull/14654) or [our llama.cpp fork](https://github.com/unslothai/llama.cpp) (easier to work)
+- For complete detailed instructions, see our guide: [docs.unsloth.ai/basics/kimi-k2](https://docs.unsloth.ai/basics/kimi-k2)
 
 <div align="center">
   <picture>