vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

add tests for /props and /slots

Browse source
This commit is contained in:
Xuan Son Nguyen 2024-12-07 19:17:35 +01:00
parent 6bf6e3066c
commit e721f4c6b4
2 changed files with 22 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

20
examples/server/tests/unit/test_basic.py
View file

@ -23,6 +23,10 @@ def test_server_props():
res = server.make_request("GET", "/props") res = server.make_request("GET", "/props")
assert res.status_code == 200 assert res.status_code == 200
assert res.body["total_slots"] == server.n_slots 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(): def test_server_models():
@ -36,12 +40,26 @@ def test_server_models():
def test_server_slots(): def test_server_slots():
global server 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.server_slots = True
server.n_slots = 2
server.start() server.start()
res = server.make_request("GET", "/slots") res = server.make_request("GET", "/slots")
assert res.status_code == 200 assert res.status_code == 200
assert len(res.body) == server.n_slots assert len(res.body) == server.n_slots
assert res.body[0]["n_ctx"] > 0 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" in res.body[0]
assert res.body[0]["params"]["seed"] == server.seed
def test_load_split_model(): def test_load_split_model():

7
examples/server/tests/utils.py
View file

@ -92,7 +92,6 @@ class ServerProcess:
else: else:
server_path = "../../../build/bin/llama-server" server_path = "../../../build/bin/llama-server"
server_args = [ server_args = [
"--slots", # requires to get slot status via /slots endpoint
"--host", "--host",
self.server_host, self.server_host,
"--port", "--port",
@ -184,7 +183,7 @@ class ServerProcess:
start_time = time.time() start_time = time.time()
while time.time() - start_time < timeout_seconds: while time.time() - start_time < timeout_seconds:
try: try:
response = self.make_request("GET", "/slots", headers={ response = self.make_request("GET", "/health", headers={
"Authorization": f"Bearer {self.api_key}" if self.api_key else None "Authorization": f"Bearer {self.api_key}" if self.api_key else None
}) })
if response.status_code == 200: if response.status_code == 200:
@ -227,7 +226,7 @@ class ServerProcess:
result.headers = dict(response.headers) result.headers = dict(response.headers)
result.status_code = response.status_code result.status_code = response.status_code
result.body = response.json() if parse_body else None result.body = response.json() if parse_body else None
print("Response from server", result.body) print("Response from server", json.dumps(result.body, indent=2))
return result return result
def make_stream_request( def make_stream_request(
@ -248,7 +247,7 @@ class ServerProcess:
break break
elif line.startswith('data: '): elif line.startswith('data: '):
data = json.loads(line[6:]) data = json.loads(line[6:])
print("Partial response from server", data) print("Partial response from server", json.dumps(data, indent=2))
yield data yield data