Skip to content

Instantly share code, notes, and snippets.

View juangon's full-sized avatar

Juan G. juangon

  • Valladolid, Spain.
View GitHub Profile
@juangon
juangon / grpo_demo.py
Created January 31, 2025 10:40 — forked from willccbb/grpo_demo.py
GRPO Llama-1B
# train_grpo.py
import re
import torch
from datasets import load_dataset, Dataset
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import LoraConfig
from trl import GRPOConfig, GRPOTrainer
# Load and prep dataset