youngkingdom/vllm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[Model] Support Qwen2 embeddings and use tags to select model tests (#10184)

Browse Source
This commit is contained in:
Cyrus Leung
2024-11-15 12:23:09 +08:00
committed by GitHub
parent 2885ba0e24
commit b40cf6402e
19 changed files with 252 additions and 178 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
tests/models/encoder_decoder/language/test_bart.py
View File

@ -14,8 +14,6 @@ from ....conftest import (DecoderPromptType, ExplicitEncoderDecoderPrompt,
from ....utils import multi_gpu_test
from ...utils import check_logprobs_close
MODELS = ["facebook/bart-base", "facebook/bart-large-cnn"]
def vllm_to_hf_output(
vllm_output: Tuple[List[int], str, Optional[SampleLogprobs]],
@ -170,7 +168,14 @@ def run_test(
)
@pytest.mark.parametrize("model", MODELS)
@pytest.mark.parametrize(
"model",
[
pytest.param("facebook/bart-base",
marks=[pytest.mark.core_model, pytest.mark.cpu_model]),
pytest.param("facebook/bart-large-cnn"),
],
)
@pytest.mark.parametrize("dtype", ["float", "bfloat16"])
@pytest.mark.parametrize("max_tokens", [64])
@pytest.mark.parametrize("num_logprobs", [5])

3
tests/models/encoder_decoder/vision_language/test_mllama.py
View File

@ -233,6 +233,7 @@ def clear_cache():
@large_gpu_test(min_gb=48)
@pytest.mark.core_model
@pytest.mark.parametrize("model", models)
@pytest.mark.parametrize(
"sizes",
@ -278,6 +279,7 @@ def test_models_single_leading_image(hf_runner, vllm_runner, image_assets,
@large_gpu_test(min_gb=48)
@pytest.mark.core_model
@pytest.mark.parametrize("model", models)
@pytest.mark.parametrize("dtype", ["bfloat16"])
@pytest.mark.parametrize("max_tokens", [128])
@ -326,6 +328,7 @@ def test_models_multi_leading_images(hf_runner, vllm_runner, image_assets,
@large_gpu_test(min_gb=48)
@pytest.mark.core_model
@pytest.mark.parametrize("model", models)
@pytest.mark.parametrize("dtype", ["bfloat16"])
@pytest.mark.parametrize("max_tokens", [128])