File size: 2,732 Bytes
db56fd6
 
 
 
 
 
 
 
 
 
 
 
 
 
228cd9c
db56fd6
228cd9c
db56fd6
 
 
228cd9c
db56fd6
 
228cd9c
db56fd6
 
 
 
 
228cd9c
db56fd6
 
 
228cd9c
 
 
 
 
 
 
db56fd6
228cd9c
db56fd6
 
 
 
 
 
 
228cd9c
 
 
db56fd6
228cd9c
 
 
 
 
 
 
 
 
 
 
 
db56fd6
228cd9c
 
 
 
 
 
 
db56fd6
 
 
 
228cd9c
 
 
 
db56fd6
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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
# Copyright 2025 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import os
import requests
from typing import Optional, Union, List
from auth import create_credentials, get_access_token_refresh_if_needed
from cache import cache

_endpoint_url = os.environ.get("GCP_MEDGEMMA_ENDPOINT")

# Create credentials
secret_key_json = os.environ.get("GCP_MEDGEMMA_SERVICE_ACCOUNT_KEY")
medgemma_credentials = create_credentials(secret_key_json)

# https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints.chat/completions
@cache.memoize()
def medgemma_get_text_response(
    messages: List[dict],
    temperature: float = 0.1,
    max_tokens: int = 4096,
    stream: bool = False,
    top_p: Optional[float] = None,
    seed: Optional[int] = None,
    stop: Optional[Union[List[str], str]] = None,
    frequency_penalty: Optional[float] = None,
    presence_penalty: Optional[float] = None,
    model: str = "tgi"
) -> str:
    """
    Makes a chat completion request to the configured LLM API (Vertex AI/OpenAI-compatible).
    """
    headers = {
        "Authorization": f"Bearer {get_access_token_refresh_if_needed(medgemma_credentials)}",
        "Content-Type": "application/json",
    }

    payload = {
        "messages": messages,
        "max_tokens": max_tokens
    }

    if temperature is not None:
        payload["temperature"] = temperature
    if top_p is not None:
        payload["top_p"] = top_p
    if seed is not None:
        payload["seed"] = seed
    if stop is not None:
        payload["stop"] = stop
    if frequency_penalty is not None:
        payload["frequency_penalty"] = frequency_penalty
    if presence_penalty is not None:
        payload["presence_penalty"] = presence_penalty

    response = requests.post(
        _endpoint_url,
        headers=headers,
        json=payload,
        stream=stream,
        timeout=60
    )
    try:
        response.raise_for_status()
        return response.json()["choices"][0]["message"]["content"]
    except requests.exceptions.JSONDecodeError:
        print(
            f"Error: Failed to decode JSON from MedGemma. "
            f"Status: {response.status_code}, Response: {response.text}"
        )
        raise