Any starting guide for beginners?
Hi guys so recently I got into this stuff playing with various local llm models and trying out mcp's and other interesting stuff like heretic versions and all.
However as I am diving deep, I am not able to understand a lot of terms like top p, k, inference, training, llama cpp, presence, repeat penalty, gguf and hardware thing like layers, universal kv etc etc which I want to tune as per my liking or atleast understand how they effect the responses.
Now I am a complete newcomer to this stuff, not in AI like rag and stuff but more towards learning and understanding this terminology. Is there any guide someone can provide which can help me understand this stuff? I know this may be a part of ML or deep learning but I may not want to go that route and learning in that domain purely but just enough so I can play with models that community provides on huggingface for instance. Just enough knowledge so I can understand what author has changed, and what diff diff terminology or methods or this llama, vllm stuff means and what are all these overwhelmimg settings on lm/unsloth studio.
Thanks in advance.