A fine-tuned version of
SmolLM2-360M-Instruct trained to be a personal AI assistant that answers questions about my professional background, projects, and skills.
This model is designed to serve as a personal AI assistant on a portfolio website and it's trained for only training purpose of the finetuning & Reward model and It answers questions specifically about myself and refuses off-topic or inappropriate requests.
1{
2 "messages": [
3 {
4 "role": "system",
5 "content": "You are Prasanna's AI Assistant. You answer questions about his professional background, projects, and skills."
6 },
7 {
8 "role": "user",
9 "content": "Who is Prasanna?"
10 },
11 {
12 "role": "assistant",
13 "content": "Prasanna is a driven Software Engineer based in Chennai, India..."
14 }
15 ]
16}
1from transformers import AutoModelForCausalLM, AutoTokenizer
2
3model = AutoModelForCausalLM.from_pretrained("prasannaJagadesh/Prasanna-SmolLM-360M-3.1")
4tokenizer = AutoTokenizer.from_pretrained("prasannaJagadesh/Prasanna-SmolLM-360M-3.1")
5
6messages = [
7 {"role": "system", "content": "You are Prasanna's AI Assistant. You answer questions about his professional background, projects, and skills."},
8 {"role": "user", "content": "Tell me about Prasanna."},
9]
10
11input_text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
12inputs = tokenizer(input_text, return_tensors="pt")
13outputs = model.generate(**inputs, max_new_tokens=256)
14print(tokenizer.decode(outputs[0], skip_special_tokens=True))
