Interactive Code Playground — AI Tutorial Examples
Reviewed: June 4, 2026
Try the code examples from our AI tutorials directly in your browser. Each example is editable and runnable via linked Colab notebooks or local setup instructions.
Available Examples
🤖 Example 1: Basic AI Chatbot
Build a conversational chatbot with memory and streaming responses. Uses LangChain + OpenAI/Ollama.
from langchain_openai import ChatOpenAI
from langchain_core.messages import HumanMessage, AIMessage, SystemMessage
from langchain_core.chat_history import InMemoryChatMessageHistory
llm = ChatOpenAI(model="gpt-4o-mini", streaming=True)
history = InMemoryChatMessageHistory()
history.add_message(SystemMessage(content="You are a helpful assistant."))
def chat(text):
history.add_message(HumanMessage(content=text))
response = ""
for chunk in llm.stream(history.messages):
print(chunk.content or "", end="")
response += chunk.content or ""
history.add_message(AIMessage(content=response))
return response
chat("Hello! What can you help me with?")
🧠 Example 2: RAG Pipeline
Retrieve relevant documents and feed them to an LLM for grounded responses.
from langchain_community.vectorstores import Chroma
from langchain_openai import OpenAIEmbeddings, ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.runnables import RunnablePassthrough
# Build vector store
embeddings = OpenAIEmbeddings()
vectorstore = Chroma.from_texts(
["Document 1 content...", "Document 2 content..."],
embeddings
)
retriever = vectorstore.as_retriever()
# RAG chain
prompt = ChatPromptTemplate.from_template(
"Answer based on context:n{context}nnQuestion: {question}"
)
llm = ChatOpenAI(model="gpt-4o-mini")
def format_docs(docs):
return "nn".join(d.page_content for d in docs)
rag_chain = (
{"context": retriever | format_docs, "question": RunnablePassthrough()}
| prompt | llm
)
result = rag_chain.invoke("What is the main topic?")
print(result.content)
🎯 Example 3: LoRA Fine-Tuning Loop
The core training loop for fine-tuning a model with LoRA adapters.
from unsloth import FastLanguageModel
from trl import SFTTrainer
from transformers import TrainingArguments
import torch
# Load model
model, tokenizer = FastLanguageModel.from_pretrained(
"unsloth/Llama-3.2-1B-Instruct",
max_seq_length=2048,
load_in_4bit=True,
)
# Apply LoRA
model = FastLanguageModel.get_peft_model(
model, r=16,
target_modules=["q_proj","k_proj","v_proj","o_proj",
"gate_proj","up_proj","down_proj"],
)
# Train
trainer = SFTTrainer(
model=model,
tokenizer=tokenizer,
train_dataset=dataset["train"],
dataset_text_field="text",
max_seq_length=2048,
args=TrainingArguments(
per_device_train_batch_size=2,
gradient_accumulation_steps=4,
num_train_epochs=3,
learning_rate=2e-4,
output_dir="outputs",
optim="adamw_8bit",
),
)
trainer.train()
# Save
model.save_pretrained_merged("my_finetuned_model",
tokenizer, save_method="merged_16bit")
print("Training complete! Model saved.")
📊 Example 4: Evaluating Model Output
Simple evaluation harness for measuring response quality.
from langchain.evaluation import load_evaluator
from langchain_openai import ChatOpenAI
evaluator = load_evaluator(
"criteria",
criteria="conciseness",
llm=ChatOpenAI(model="gpt-4o-mini")
)
prediction = "The capital of France, which is a country in Western Europe, is Paris."
reference = "Paris"
result = evaluator.evaluate_strings(
prediction=prediction,
reference=reference,
)
print(f"Score: {result['score']}")
print(f"Reasoning: {result['reasoning']}")
Running Locally
To run these examples on your own machine:
git clone https://github.com/datagate-ch/ai-tutorials.git
cd ai-tutorials
pip install -r requirements.txt
# Copy .env.example and fill in your API keys
cp .env.example .env
# Run any example
python examples/01_basic_chatbot.py
Tips
- Most examples work with both OpenAI and local Ollama models — just change the model name
- Colab free tier provides a T4 GPU, sufficient for inference and small fine-tuning runs
- For larger fine-tuning, use a cloud GPU (Lambda, Modal, or Vast.ai)
Related tutorials: Build Your First AI Chatbot · Fine-Tune an LLM on Custom Data