Sandeep GrubBox

C, go, zig, Odin, jai, mojo, Assembly, python,swift. Neural Network, Blockchain, Arduino (wifi jammer), Bash/fish/Zsh, Nvim/Vscode, i3/sway, Unix

willccbb / grpo_demo.py

Last active May 9, 2025 03:34

GRPO Llama-1B

	# train_grpo.py
	#
	# See https://github.com/willccbb/verifiers for ongoing developments
	#
	import re
	import torch
	from datasets import load_dataset, Dataset
	from transformers import AutoTokenizer, AutoModelForCausalLM
	from peft import LoraConfig
	from trl import GRPOConfig, GRPOTrainer