willccbb

# train_grpo.py
#
# See https://github.com/willccbb/verifiers for ongoing developments
#
"""
citation:

@misc{brown2025grpodemo,
  title={Granular Format Rewards for Eliciting Mathematical Reasoning Capabilities in Small Language Models},
  author={Brown, William},

vgel

import argparse
import random
import sys

from transformers import AutoModelForCausalLM, AutoTokenizer, DynamicCache
import torch

parser = argparse.ArgumentParser()
parser.add_argument("question", type=str)
parser.add_argument(

balouf

ed1d1a8d

import numpy as np
import openai
import scipy.special

import tiktoken


def get_top_chat_logprobs(
    model: str,
    messages: list[dict[str, str]],

sientifiko

library(tidyverse)
theme_set(theme_bw(base_size = 21))
options(scipen = 999)

dat <- read.csv("dataregionesChile.csv")

# filtrar el año 2020
y2020 <- dat %>% filter(year == 2020)

# generar histograma

jbwhit

RaphaelWimmer

#!/usr/bin/env python3 
import openai
import time
import matplotlib.pyplot as plt

def pick_numbers(n, model, temperature, clean_session):
    numbers = []
    messages = []
    system_msg = "Please only respond with the number, don't say anything else."
    messages.append({"role": "system", "content": system_msg})

kylemcdonald

tcapelle

import os
import openai
from rich.console import Console

console = Console()

openai.api_key = os.getenv("OPENAI_API_KEY")
    
history = [{"role": "system", "content": "You are a helpful assistant."},]

AustinRochford