| from transformers import AutoImageProcessor, Swinv2Model | |
| import torch | |
| device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') | |
| class EndpointHandler(): | |
| def __init__(self, path=""): | |
| self.model = Swinv2Model.from_pretrained("microsoft/moooji/swinv2-large-patch4-window12to24-192to384-22kto1k-ft").to(device) | |
| self.processor = AutoImageProcessor.from_pretrained("microsoft/moooji/swinv2-large-patch4-window12to24-192to384-22kto1k-ft") | |
| def __call__(self, data: Any) -> List[float]: | |
| inputs = data.pop("inputs", data) | |
| image = Image.open(BytesIO(base64.b64decode(inputs['image']))) | |
| inputs = self.processor(image, return_tensors="pt").to(device) | |
| with torch.no_grad(): | |
| outputs = model(**inputs) | |
| last_hidden_states = outputs.last_hidden_state | |
| return last_hidden_states[2].tolist() | |