Skip to content

Instantly share code, notes, and snippets.

View huseinzol05's full-sized avatar
🍵
Hunting cendol!

HUSEIN ZOLKEPLI huseinzol05

🍵
Hunting cendol!
View GitHub Profile
@huseinzol05
huseinzol05 / grpo_demo.py
Created January 30, 2025 22:27 — forked from willccbb/grpo_demo.py
GRPO Llama-1B LOL
# 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