Skip to content

Instantly share code, notes, and snippets.

View rbtrsv's full-sized avatar

Robert Radoslav rbtrsv

  • London, United Kingdom
View GitHub Profile
@rbtrsv
rbtrsv / grpo_demo.py
Created January 31, 2025 15:13 — 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
class UploadPage extends StatefulWidget {
UploadPage({Key key, this.url}) : super(key: key);
final String url;
@override
_UploadPageState createState() => _UploadPageState();
}
class _UploadPageState extends State<UploadPage> {