app.py

import gradio as gr
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("georgiyozhegov/calculator-8m")
model = AutoModelForCausalLM.from_pretrained("georgiyozhegov/calculator-8m")

def solve(problem):
    prompt = f"find {problem}\nstep"
    inputs = tokenizer(prompt, return_tensors="pt", return_token_type_ids=False)

    with torch.no_grad():
        outputs = model.generate(
            input_ids=inputs["input_ids"],
            attention_mask=inputs["attention_mask"],
            max_length=32,
            do_sample=True,
            top_k=50,
            top_p=0.98
        )

    count = 0
    for index, token in enumerate(outputs[0]):
        if token == 6: count += 1
        if count >= 2: break

    output = tokenizer.decode(outputs[0][:index])
    return output

examples = [
    ["2 + 3"],
    ["10 / 0.5"],
]

demo = gr.Interface(
    fn=solve,
    inputs=gr.Textbox(lines=5, label="Problem"),
    outputs=gr.Textbox(label="Solution"),
    examples=examples,
)

demo.launch()