test_lazy_outlines.py 2.71 KB
Newer Older
1
2
# SPDX-License-Identifier: Apache-2.0

3
import sys
youkaichao's avatar
youkaichao committed
4
from contextlib import nullcontext
5

youkaichao's avatar
youkaichao committed
6
from vllm_test_utils import BlameResult, blame
youkaichao's avatar
youkaichao committed
7

8
from vllm import LLM, SamplingParams
9
from vllm.distributed import cleanup_dist_env_and_memory
10
11


youkaichao's avatar
youkaichao committed
12
def run_normal():
13
14
15
16
17
18
19
20
    prompts = [
        "Hello, my name is",
        "The president of the United States is",
        "The capital of France is",
        "The future of AI is",
    ]
    sampling_params = SamplingParams(temperature=0.8, top_p=0.95)

21
    # Create an LLM without guided decoding as a baseline.
22
23
24
25
26
27
28
29
30
    llm = LLM(model="facebook/opt-125m",
              enforce_eager=True,
              gpu_memory_utilization=0.3)
    outputs = llm.generate(prompts, sampling_params)
    for output in outputs:
        prompt = output.prompt
        generated_text = output.outputs[0].text
        print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")

31
32
33
34
    # Destroy the LLM object and free up the GPU memory.
    del llm
    cleanup_dist_env_and_memory()

youkaichao's avatar
youkaichao committed
35
36

def run_lmfe(sample_regex):
37
    # Create an LLM with guided decoding enabled.
38
39
40
    llm = LLM(model="facebook/opt-125m",
              enforce_eager=True,
              guided_decoding_backend="lm-format-enforcer",
41
              gpu_memory_utilization=0.3)
42
43
44
45
46
47
48
49
50
51
52
53
54
55
    sampling_params = SamplingParams(temperature=0.8, top_p=0.95)
    outputs = llm.generate(
        prompts=[
            f"Give an example IPv4 address with this regex: {sample_regex}"
        ] * 2,
        sampling_params=sampling_params,
        use_tqdm=True,
        guided_options_request=dict(guided_regex=sample_regex))

    for output in outputs:
        prompt = output.prompt
        generated_text = output.outputs[0].text
        print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")

youkaichao's avatar
youkaichao committed
56
57
58
59

def test_lazy_outlines(sample_regex):
    """If users don't use guided decoding, outlines should not be imported.
    """
60
    # make sure outlines is not imported
youkaichao's avatar
youkaichao committed
61
    module_name = "outlines"
youkaichao's avatar
youkaichao committed
62
63
64
65
66
67
68
69
70
    # In CI, we only check finally if the module is imported.
    # If it is indeed imported, we can rerun the test with `use_blame=True`,
    # which will trace every function call to find the first import location,
    # and help find the root cause.
    # We don't run it in CI by default because it is slow.
    use_blame = False
    context = blame(
        lambda: module_name in sys.modules) if use_blame else nullcontext()
    with context as result:
youkaichao's avatar
youkaichao committed
71
72
        run_normal()
        run_lmfe(sample_regex)
youkaichao's avatar
youkaichao committed
73
74
75
76
77
78
    if use_blame:
        assert isinstance(result, BlameResult)
        print(f"the first import location is:\n{result.trace_stack}")
    assert module_name not in sys.modules, (
        f"Module {module_name} is imported. To see the first"
        f" import location, run the test with `use_blame=True`.")