!pip install -q "torchao>=0.16.0"

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel
import torch

tokenizer = AutoTokenizer.from_pretrained("jpo89/qwen2.5-3b-sft-dapo")

model = AutoModelForCausalLM.from_pretrained(
    "Qwen/Qwen2.5-3B-Instruct",
    dtype=torch.bfloat16,
    device_map="auto",
)

model = PeftModel.from_pretrained(model, "jpo89/qwen2.5-3b-sft-dapo")
model.eval()
print("Model ready.")

Evaluation

Math Reasoning Benchmark Suite Dashboard

The qwen2.5-3b-sft-dapo model was evaluated across a 3-dataset math reasoning benchmark suite (n=15 per dataset), demonstrating robust analytical capabilities. As visualized in the interactive performance dashboard (Unknown-2.png), the model achieves an impressive 86.7% accuracy on grade-school math (GSM8K), 60.0% on math word problems (MAWPS), and 53.3% on arithmetic reasoning (ASDiv). The cumulative accuracy curves highlight rapid stabilization and consistent execution well above the 50% baseline, positioning this 3B parameter model as a highly efficient and competitive asset for complex algorithmic and logical tasks.

Test the model (sample testing)

python

def test(question):
    messages = [{"role": "user", "content": question}]
    inputs = tokenizer.apply_chat_template(
        messages,
        add_generation_prompt=True,
        return_tensors="pt",
        return_dict=True,
    ).to(model.device)

    with torch.no_grad():
        output = model.generate(
            **inputs,
            max_new_tokens=256,
            do_sample=True,
            temperature=0.4,
            top_p=0.9,
            repetition_penalty=1.1,
            pad_token_id=tokenizer.eos_token_id,
        )

    response = tokenizer.decode(
        output[0][inputs["input_ids"].shape[-1]:],
        skip_special_tokens=True
    )
    print(f"Q: {question}")
    print(f"A: {response}")
    print("-" * 60)

# Test with a few questions
test("If x + 5 = 12, what is x?")
test("What is the sum of the first 10 natural numbers?")
test("A train travels 60 km/h for 2.5 hours. How far does it travel?")

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

Hardware Type: [More Information Needed]
Hours used: [More Information Needed]
Cloud Provider: [More Information Needed]
Compute Region: [More Information Needed]
Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Framework versions

PEFT 0.19.1