Harshit-Makraria/the-pivot-advisor-3b

The Pivot — Co-Founder Advisor (Qwen2.5-3B, LoRA)

Conversational co-founder advisor fine-tuned with GRPO on the The Pivot OpenEnv environment.

• Persona: Alex, a no-nonsense co-founder advisor with 15+ years of startup experience.
• Output style: 2-3 sentence natural-language replies (not action labels).
• Memory: 3-turn rolling conversation memory.
• Trained on: 100 episodes × 20 steps across 5 startup scenarios.

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

base = AutoModelForCausalLM.from_pretrained('Qwen/Qwen2.5-3B-Instruct')
model = PeftModel.from_pretrained(base, 'Harshit-Makraria/the-pivot-advisor-3b')
tokenizer = AutoTokenizer.from_pretrained('Harshit-Makraria/the-pivot-advisor-3b')

Built for the Meta PyTorch OpenEnv Hackathon 2026.