How do I install the Large Language Models skill?

Run npx skillmds@latest add ffsshhttiikk/large-language-models in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Large Language Models skill do?

Large language models It is listed under Coding & Dev Tools on SkillMD.

Is Large Language Models safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Large Language Models?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Large Language Models free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.

Who published Large Language Models?

ffsshhttiikk (@ffsshhttiikk) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Large Language Models by ffsshhttiikk

Name: Large Language Models
Author: ffsshhttiikk

File contents

What I do

Work with large language models
Implement prompt engineering
Fine-tune LLM models
Build RAG systems
Optimize LLM inference
Handle LLM hallucination

When to use me

Use me when:

Building LLM-powered applications
Implementing chatbots
Creating text generation systems
Knowledge-intensive question answering
Text classification with LLMs

Key Concepts

LLM Architecture

Transformer: Self-attention based
Encoder-only: BERT, RoBERTa
Decoder-only: GPT, LLaMA, Mistral
Encoder-decoder: T5, BART

Prompt Engineering

# Few-shot prompting
prompt = """Classify the sentiment as POSITIVE, NEGATIVE, or NEUTRAL.

Text: This product is amazing!
Sentiment: POSITIVE

Text: Terrible experience, would not recommend.
Sentiment: NEGATIVE

Text: It was okay, nothing special.
Sentiment: NEUTRAL

Text: I absolutely love this!
Sentiment:"""

# Chain-of-thought
cot_prompt = """Solve the problem step by step.

Problem: If there are 5 birds and you shoot 2, how many remain?
Solution: First, 5 - 2 = 3. Then, the remaining birds fly away because of the noise, so 0 remain. Answer: 0

Problem: John has 3 apples. He buys 5 more. Then he eats 2. How many?
Solution:"""

Fine-tuning

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import LoraConfig, get_peft_model, TaskType

# LoRA fine-tuning
model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-2-7b")
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b")

lora_config = LoraConfig(
    r=16,
    lora_alpha=32,
    target_modules=["q_proj", "v_proj"],
    lora_dropout=0.05,
    bias="none",
    task_type=TaskType.CAUSAL_LM
)

model = get_peft_model(model, lora_config)
# Fine-tune on custom dataset

RAG (Retrieval-Augmented Generation)

from langchain.vectorstores import Chroma
from langchain.embeddings import OpenAIEmbeddings
from langchain.chains import RetrievalQA
from langchain.llms import OpenAI

# Build retriever
embeddings = OpenAIEmbeddings()
vectorstore = Chroma.from_documents(documents, embeddings)
retriever = vectorstore.as_retriever()

# Create QA chain
qa = RetrievalQA.from_chain_type(
    llm=OpenAI(),
    chain_type="stuff",
    retriever=retriever
)

answer = qa.run(query)

Challenges

Hallucination
Context length limits
Token costs
Latency
Bias and safety

---name: large-language-modelsdescription: Large language modelslicense: MIT---## What I do- Work with large language models- Implement prompt engineering- Fine-tune LLM models- Build RAG systems- Optimize LLM inference- Handle LLM hallucination## When to use meUse me when:- Building LLM-powered applications- Implementing chatbots- Creating text generation systems- Knowledge-intensive question answering- Text classification with LLMs## Key Concepts### LLM Architecture- **Transformer**: Self-attention based- **Encoder-only**: BERT, RoBERTa- **Decoder-only**: GPT, LLaMA, Mistral- **Encoder-decoder**: T5, BART### Prompt Engineering```python# Few-shot promptingprompt = """Classify the sentiment as POSITIVE, NEGATIVE, or NEUTRAL.Text: This product is amazing!Sentiment: POSITIVEText: Terrible experience, would not recommend.Sentiment: NEGATIVEText: It was okay, nothing special.Sentiment: NEUTRALText: I absolutely love this!Sentiment:"""# Chain-of-thoughtcot_prompt = """Solve the problem step by step.Problem: If there are 5 birds and you shoot 2, how many remain?Solution: First, 5 - 2 = 3. Then, the remaining birds fly away because of the noise, so 0 remain. Answer: 0Problem: John has 3 apples. He buys 5 more. Then he eats 2. How many?Solution:"""```### Fine-tuning```pythonfrom transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import LoraConfig, get_peft_model, TaskType# LoRA fine-tuningmodel = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-2-7b")tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b")lora_config = LoraConfig( r=16, lora_alpha=32, target_modules=["q_proj", "v_proj"], lora_dropout=0.05, bias="none", task_type=TaskType.CAUSAL_LM)model = get_peft_model(model, lora_config)# Fine-tune on custom dataset```### RAG (Retrieval-Augmented Generation)```pythonfrom langchain.vectorstores import Chromafrom langchain.embeddings import OpenAIEmbeddingsfrom langchain.chains import RetrievalQAfrom langchain.llms import OpenAI# Build retrieverembeddings = OpenAIEmbeddings()vectorstore = Chroma.from_documents(documents, embeddings)retriever = vectorstore.as_retriever()# Create QA chainqa = RetrievalQA.from_chain_type( llm=OpenAI(), chain_type="stuff", retriever=retriever)answer = qa.run(query)```### Challenges- Hallucination- Context length limits- Token costs- Latency- Bias and safety

Frequently asked questions

Large Language Models