|
from typing import Dict, List, Any |
|
from transformers import pipeline |
|
from PIL import Image |
|
import requests |
|
|
|
class EndpointHandler(): |
|
def __init__(self, path=""): |
|
model_id = "llava-hf/llava-interleave-qwen-0.5b-hf" |
|
self.pipe = pipeline("image-to-text", model=model_id) |
|
|
|
def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]: |
|
""" |
|
data args: |
|
inputs (:obj: `str` | `PIL.Image` | `np.array`) |
|
kwargs |
|
Return: |
|
A :obj:`list` | `dict`: will be serialized and returned |
|
""" |
|
|
|
inputs = data.pop('inputs', data) |
|
url = inputs.get('url') |
|
prompt = inputs.get('prompt') |
|
max_new_tokens = inputs.get('max_new_tokens', 1000) |
|
|
|
image = Image.open(requests.get(url, stream=True).raw) |
|
prompt = f'user<image>\n{prompt}\nassistant:' |
|
|
|
results = self.pipe(image, prompt=prompt, generate_kwargs={"max_new_tokens": max_new_tokens}) |
|
result = results[0] |
|
|
|
return result |
