File size: 1,504 Bytes
66de2f3
 
 
 
 
6f9e0c8
66de2f3
ed463ac
66de2f3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3081d6b
 
358c97c
3081d6b
 
a364e6d
3081d6b
 
ed463ac
66de2f3
 
 
 
 
 
 
 
 
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
42
43
44
from typing import Dict, List, Any
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, pipeline
import torch

class EndpointHandler:
    def __init__(self, path=""):
        # Load model and processor from path
        self.model = AutoModelForSeq2SeqLM.from_pretrained(path)
        self.tokenizer = AutoTokenizer.from_pretrained(path)

    def __call__(self, data: Dict[str, Any]) -> Dict[str, str]:
        """
        Args:
            data (:obj:):
                Includes the deserialized image file as PIL.Image
        """
        # Process input
        inputs = data.pop("inputs", data)
        parameters = data.pop("parameters", None)

        # Preprocess
        input_ids = self.tokenizer(inputs, return_tensors="pt").input_ids

        # Modify parameters to increase max_length
        if parameters is None:
            parameters = {}
        parameters['max_length'] = 1012  # Set your desired max_length here
        parameters['min_length'] = 100
        parameters['length_penalty'] = 10.0
        parameters['num_beams'] = 25
        parameters['early_stopping'] = True
        parameters['temperature'] = 0.5
        parameters['top_k'] = 25
        parameters['top_p'] = 1.0
        
        

        # Generate output
        outputs = self.model.generate(input_ids, **parameters)
        
        # Postprocess the prediction
        prediction = self.tokenizer.decode(outputs[0], skip_special_tokens=True)
        
        return [{"generated_text": prediction}]