frontend.py 2.02 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# 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 logging
import signal
import sys

from components.processor import Processor
from components.utils import GeneralRequest
22
23
from fastapi import FastAPI
from fastapi.responses import StreamingResponse
24

25
from dynamo.sdk import depends, dynamo_endpoint, service
26
27
28
29
from dynamo.sdk.lib.image import DYNAMO_IMAGE

logger = logging.getLogger(__name__)

30
31
app = FastAPI(title="Hello World LLM")

32
33

@service(
34
    dynamo={"enabled": True, "namespace": "dynamo-demo"},
35
    image=DYNAMO_IMAGE,
36
    app=app,
37
38
39
40
41
42
43
44
45
46
47
48
)
class Frontend:
    processor = depends(Processor)

    def __init__(self):
        signal.signal(signal.SIGTERM, self.handle_exit)
        signal.signal(signal.SIGINT, self.handle_exit)

    def handle_exit(self, signum, frame):
        logger.debug(f"Received signal {signum}, shutting down...")
        sys.exit(0)

49
    @dynamo_endpoint(is_api=True)
50
51
52
    async def generate(self, prompt, request_id):  # from request body keys
        """Stream results from the pipeline."""
        logger.info(f"Received: {prompt=},{request_id=}")
53
54
55
56
57
58
59
60
61

        async def content_generator():
            frontend_request = GeneralRequest(
                prompt=prompt, request_id=request_id
            ).model_dump_json()
            async for response in self.processor.processor_generate(frontend_request):
                yield f"Response: {response}\n"

        return StreamingResponse(content_generator())