Completed Exercise 1 - Hello LLM Endpoint

done.txt

@@ -0,0 +1,5 @@
+NameName: Delsy Kinyuy
+Exercise: 1 - Hello LLM Endpoint
+Challenges: 
+- Installing transformers took some time
+- Model download was large on first run

exercise1.py

@@ -0,0 +1,30 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+from transformers import pipeline
+
+# init FastAPI
+app = FastAPI(title="Generative AI Exercises")
+
+# load Hugging Face model (distilgpt2 is small enough to run on CPU)
+generator = pipeline("text-generation", model="distilgpt2")
+
+# request schema
+class PromptRequest(BaseModel):
+    prompt: str
+    max_tokens: int = 100
+
+@app.get("/")
+def root():
+    return {"message": "FastAPI + Hugging Face is live 🚀"}
+
+@app.post("/hello-llm")
+def hello_llm(request: PromptRequest):
+    """
+    Generate text from a given prompt using distilgpt2.
+    """
+    output = generator(
+        request.prompt,
+        max_length=len(request.prompt.split()) + request.max_tokens,
+        num_return_sequences=1
+    )
+    return {"prompt": request.prompt, "generated": output[0]["generated_text"]}