File size: 2,877 Bytes
4d35814 | 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 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 | import pytest
import requests
from utils import *
server = ServerPreset.tinyllama2()
@pytest.fixture(scope="session", autouse=True)
def do_something():
# this will be run once per test session, before any tests
ServerPreset.load_all()
@pytest.fixture(autouse=True)
def create_server():
global server
server = ServerPreset.tinyllama2()
def test_server_start_simple():
global server
server.start()
res = server.make_request("GET", "/health")
assert res.status_code == 200
def test_server_props():
global server
server.start()
res = server.make_request("GET", "/props")
assert res.status_code == 200
assert ".gguf" in res.body["model_path"]
assert res.body["total_slots"] == server.n_slots
default_val = res.body["default_generation_settings"]
assert server.n_ctx is not None and server.n_slots is not None
assert default_val["n_ctx"] == server.n_ctx / server.n_slots
assert default_val["params"]["seed"] == server.seed
def test_server_models():
global server
server.start()
res = server.make_request("GET", "/models")
assert res.status_code == 200
assert len(res.body["data"]) == 1
assert res.body["data"][0]["id"] == server.model_alias
def test_server_slots():
global server
# without slots endpoint enabled, this should return error
server.server_slots = False
server.start()
res = server.make_request("GET", "/slots")
assert res.status_code == 501 # ERROR_TYPE_NOT_SUPPORTED
assert "error" in res.body
server.stop()
# with slots endpoint enabled, this should return slots info
server.server_slots = True
server.n_slots = 2
server.start()
res = server.make_request("GET", "/slots")
assert res.status_code == 200
assert len(res.body) == server.n_slots
assert server.n_ctx is not None and server.n_slots is not None
assert res.body[0]["n_ctx"] == server.n_ctx / server.n_slots
assert "params" not in res.body[0]
def test_load_split_model():
global server
server.model_hf_repo = "ggml-org/models"
server.model_hf_file = "tinyllamas/split/stories15M-q8_0-00001-of-00003.gguf"
server.model_alias = "tinyllama-split"
server.start()
res = server.make_request("POST", "/completion", data={
"n_predict": 16,
"prompt": "Hello",
"temperature": 0.0,
})
assert res.status_code == 200
assert match_regex("(little|girl)+", res.body["content"])
def test_no_webui():
global server
# default: webui enabled
server.start()
url = f"http://{server.server_host}:{server.server_port}"
res = requests.get(url)
assert res.status_code == 200
assert "<!doctype html>" in res.text
server.stop()
# with --no-webui
server.no_webui = True
server.start()
res = requests.get(url)
assert res.status_code == 404
|