File size: 1,828 Bytes
1e05993 7feaf20 6a72b4d 1e05993 6a67ddb 7feaf20 58ff319 e6b2e2d 1e05993 23be6ed 1a792e3 6a67ddb 185c874 e6b2e2d 5701230 9599168 1e05993 23be6ed 1ac6526 23be6ed 6a67ddb 1ac6526 928a2e1 1ac6526 928a2e1 b9c7f3e 928a2e1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
from typing import Dict, Any
from PIL import Image
import requests
import torch
import numpy as np
from transformers import AutoProcessor, LlavaForConditionalGeneration, BitsAndBytesConfig
class EndpointHandler():
def __init__(self, path=""):
model_id = path
self.model = LlavaForConditionalGeneration.from_pretrained(
model_id,
torch_dtype=torch.float32 ,
low_cpu_mem_usage=True,
load_in_4bit=True
)
self.processor = AutoProcessor.from_pretrained(model_id)
def __call__(self, data: Dict[str, Any]):
parameters = data.pop("inputs", data)
url = ["http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg",
"http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "USER: <image>\nWhat are these?\nASSISTANT:"
outputs = []
for link in url:
raw_image = Image.open(requests.get(link, stream=True).raw)
inputs = self.processor(prompt, raw_image, return_tensors='pt').to(0, torch.float16)
output = self.model.generate(**inputs, max_new_tokens=200, do_sample=False)
readable = (self.processor.decode(output[0][2:], skip_special_tokens=True))
outputs.append(readable)
return outputs
|