app.py

import requests
import torch
from PIL import Image

from transformers import AriaProcessor, AriaForConditionalGeneration
from fastapi import FastAPI, Request

app = FastAPI()

@app.get("/")
def greet_json():
    return {"Hello": "World!"}

@app.post("/")
async def aria_image_to_text(request: Request):
    print(1)
    data = await request.json()
    print(2)
    image_url = data.get("image_url")
    print(3)
    print('image_url')
    print(image_url)
    image = Image.open(requests.get(image_url, stream=True).raw)
    print(4)
    model_id_or_path = "rhymes-ai/Aria"
    print(5)
    model = AriaForConditionalGeneration.from_pretrained(
        model_id_or_path, device_map="auto", torch_dtype=torch.bfloat16
    )
    print(6)
    processor = AriaProcessor.from_pretrained(model_id_or_path)
    print(7)
    messages = [
        {
            "role": "user",
            "content": [
                {"type": "image"},
                {"text": "what is the image?", "type": "text"},
            ],
        }
    ]

    print(8)
    text = processor.apply_chat_template(messages, add_generation_prompt=True)
    print(9)
    inputs = processor(text=text, images=image, return_tensors="pt")
    print(10)
    inputs['pixel_values'] = inputs['pixel_values'].to(torch.bfloat16)
    print(11)
    inputs.to(model.device)
    print(12)
    output = model.generate(
        **inputs,
        max_new_tokens=15,
        stop_strings=["<|im_end|>"],
        tokenizer=processor.tokenizer,
        do_sample=True,
        temperature=0.9,
    )
    print(13)
    output_ids = output[0][inputs["input_ids"].shape[1]:]
    print(14)
    response = processor.decode(output_ids, skip_special_tokens=True)
    print(15)
    return {"response": response}

@app.get("/aria-test")
def aria_test():
    print(1)
    model_id_or_path = "rhymes-ai/Aria"
    print(2)
    model = AriaForConditionalGeneration.from_pretrained(
        model_id_or_path, device_map="auto", torch_dtype=torch.bfloat16
    )
    print(3)
    processor = AriaProcessor.from_pretrained(model_id_or_path)
    print(4)
    image = Image.open(requests.get("http://images.cocodataset.org/val2017/000000039769.jpg", stream=True).raw)
    print(5)
    messages = [
        {
            "role": "user",
            "content": [
                {"type": "image"},
                {"text": "what is the image?", "type": "text"},
            ],
        }
    ]
    print(6)    
    text = processor.apply_chat_template(messages, add_generation_prompt=True)
    print(7)
    inputs = processor(text=text, images=image, return_tensors="pt")
    print(8)
    inputs['pixel_values'] = inputs['pixel_values'].to(torch.bfloat16)
    print(9)
    inputs.to(model.device)
    print(10)
    output = model.generate(
        **inputs,
        max_new_tokens=15,
        stop_strings=["<|im_end|>"],
        tokenizer=processor.tokenizer,
        do_sample=True,
        temperature=0.9,
    )
    output_ids = output[0][inputs["input_ids"].shape[1]:]
    print(11)
    response = processor.decode(output_ids, skip_special_tokens=True)
    print(12)
    return {"response": response}