llama.cpp/examples/server/tests/unit/test_tokenize.py
Xuan Son Nguyen 45abe0f74e
server : replace behave with pytest (#10416)
* server : replace behave with pytest

* fix test on windows

* misc

* add more tests

* more tests

* styling

* log less, fix embd test

* added all sequential tests

* fix coding style

* fix save slot test

* add parallel completion test

* fix parallel test

* remove feature files

* update test docs

* no cache_prompt for some tests

* add test_cache_vs_nocache_prompt
2024-11-26 16:20:18 +01:00

60 lines
1.5 KiB
Python

import pytest
from utils import *
server = ServerPreset.tinyllama2()
@pytest.fixture(scope="module", autouse=True)
def create_server():
global server
server = ServerPreset.tinyllama2()
def test_tokenize_detokenize():
global server
server.start()
# tokenize
content = "What is the capital of France ?"
res_tok = server.make_request("POST", "/tokenize", data={
"content": content
})
assert res_tok.status_code == 200
assert len(res_tok.body["tokens"]) > 5
# detokenize
res_detok = server.make_request("POST", "/detokenize", data={
"tokens": res_tok.body["tokens"],
})
assert res_detok.status_code == 200
assert res_detok.body["content"].strip() == content
def test_tokenize_with_bos():
global server
server.start()
# tokenize
content = "What is the capital of France ?"
bosId = 1
res_tok = server.make_request("POST", "/tokenize", data={
"content": content,
"add_special": True,
})
assert res_tok.status_code == 200
assert res_tok.body["tokens"][0] == bosId
def test_tokenize_with_pieces():
global server
server.start()
# tokenize
content = "This is a test string with unicode 媽 and emoji 🤗"
res_tok = server.make_request("POST", "/tokenize", data={
"content": content,
"with_pieces": True,
})
assert res_tok.status_code == 200
for token in res_tok.body["tokens"]:
assert "id" in token
assert token["id"] > 0
assert "piece" in token
assert len(token["piece"]) > 0