[CI/Build] Move test_utils.py to tests/utils.py (#4425)

Since #4335 was merged, I've noticed that the definition of ServerRunner in the tests is the same as in the test for OpenAI API. I have moved the class to the test utilities to avoid code duplication. (Although it only has been repeated twice so far, I will add another similar test suite in #4200 which would duplicate the code a third time)

Also, I have moved the test utilities file (test_utils.py) to under the test directory (tests/utils.py), since none of its code is actually used in the main package. Note that I have added __init__.py to each test subpackage and updated the ray.init() call in the test utilities file in order to relative import tests/utils.py.
This commit is contained in:
Cyrus Leung
2024-05-13 23:50:09 +09:00
committed by GitHub
parent 702bee461f
commit 350f9e107f
34 changed files with 138 additions and 164 deletions
View File
@@ -4,10 +4,12 @@ by one. The solution is to pass arguments (model name) by environment
variables.
Run:
```sh
cd $VLLM_PATH/tests
TEST_DIST_MODEL=facebook/opt-125m pytest \
test_basic_distributed_correctness.py
distributed/test_basic_distributed_correctness.py
TEST_DIST_MODEL=meta-llama/Llama-2-7b-hf \
test_basic_distributed_correctness.py
distributed/test_basic_distributed_correctness.py
```
"""
import os
+3 -2
View File
@@ -11,8 +11,9 @@ import torch
from vllm.distributed import (broadcast_tensor_dict,
tensor_model_parallel_all_gather,
tensor_model_parallel_all_reduce)
from vllm.test_utils import (init_test_distributed_environment,
multi_process_tensor_parallel)
from ..utils import (init_test_distributed_environment,
multi_process_tensor_parallel)
@ray.remote(num_gpus=1, max_calls=1)
+3 -2
View File
@@ -10,8 +10,9 @@ from vllm.distributed.communication_op import ( # noqa
graph_capture, tensor_model_parallel_all_reduce)
from vllm.distributed.parallel_state import (get_tensor_model_parallel_group,
get_tp_ca_communicator)
from vllm.test_utils import (init_test_distributed_environment,
multi_process_tensor_parallel)
from ..utils import (init_test_distributed_environment,
multi_process_tensor_parallel)
random.seed(42)
test_sizes = [random.randint(1024, 2048 * 1024) for _ in range(8)]