Skip to content

[ci] Consolidate Qwen models #13118

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 6 commits into from
Closed

[ci] Consolidate Qwen models #13118

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
80e1d16
p
Feb 11, 2025
834270d
p
Feb 11, 2025
7b03c47
change expected val for qwen2.5
Feb 12, 2025
1b6fd7e
p
Feb 12, 2025
147e1f0
Update test_root_path.py
khluu Feb 13, 2025
e79013b
merge with main
khluu Feb 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions tests/distributed/test_pipeline_parallel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -178,6 +178,7 @@ def iter_params(self, model_id: str):
"OrionStarAI/Orion-14B-Chat": PPTestSettings.fast(),
"adept/persimmon-8b-chat": PPTestSettings.fast(),
"microsoft/phi-2": PPTestSettings.fast(),
"Qwen/Qwen2.5-1.5B-Instruct": PPTestSettings.fast(),
Copy link
Member

@DarkLight1337 DarkLight1337 Feb 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we keep this in alphabetical order (by model architecture)? Also, the test isn't actually run unless you include the model name in TEST_MODELS

"microsoft/Phi-3-small-8k-instruct": PPTestSettings.fast(),
"microsoft/Phi-3.5-MoE-instruct": PPTestSettings.detailed(multi_node_only=True, load_format="dummy"), # noqa: E501
"Qwen/Qwen-7B-Chat": PPTestSettings.fast(),
Expand Down
2 changes: 1 addition & 1 deletion tests/entrypoints/llm/test_guided_generate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@
from vllm.outputs import RequestOutput
from vllm.sampling_params import GuidedDecodingParams, SamplingParams

MODEL_NAME = "Qwen/Qwen2.5-7B-Instruct"
MODEL_NAME = "Qwen/Qwen2.5-1.5B-Instruct"
GUIDED_DECODING_BACKENDS = ["outlines", "lm-format-enforcer", "xgrammar"]


Expand Down
4 changes: 2 additions & 2 deletions tests/entrypoints/openai/correctness/test_lmeval.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,12 +15,12 @@

from ....utils import RemoteOpenAIServer

MODEL_NAME = "Qwen/Qwen2-1.5B-Instruct"
MODEL_NAME = "Qwen/Qwen2.5-1.5B-Instruct"
NUM_CONCURRENT = 500
TASK = "gsm8k"
FILTER = "exact_match,strict-match"
RTOL = 0.03
EXPECTED_VALUE = 0.58
EXPECTED_VALUE = 0.54
DEFAULT_ARGS = ["--max-model-len", "2048", "--disable-log-requests"]
MORE_ARGS_LIST = [
[], # Default
Expand Down
2 changes: 1 addition & 1 deletion tests/models/decoder_only/language/test_models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,7 +47,7 @@
marks=[pytest.mark.core_model],
),
pytest.param(
Copy link
Member

@DarkLight1337 DarkLight1337 Feb 14, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We can remove this now that #13157 has been merged.

"Qwen/Qwen-7B", # qwen (text-only)
"Qwen/Qwen2.5-1.5B", # qwen (text-only)
),
pytest.param(
"Qwen/Qwen2.5-0.5B-Instruct", # qwen2
Expand Down
2 changes: 1 addition & 1 deletion tests/prefix_caching/test_disable_sliding_window.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@

# Confirm model with sliding window works.
# config has "use_sliding_window": false
("Qwen/Qwen1.5-0.5B-Chat", 32768, 32768),
("Qwen/Qwen2.5-1.5B-Instruct", 32768, 32768),
# config has no sliding window attribute.
("TinyLlama/TinyLlama-1.1B-Chat-v1.0", 2048, 2048),
]
Expand Down
6 changes: 3 additions & 3 deletions tests/test_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ def test_incorrect_task(model_id, bad_task):


MODEL_IDS_EXPECTED = [
("Qwen/Qwen1.5-7B", 32768),
("Qwen/Qwen2.5-1.5B", 131072),
("mistralai/Mistral-7B-v0.1", 4096),
("mistralai/Mistral-7B-Instruct-v0.2", 32768),
]
Expand Down Expand Up @@ -81,9 +81,9 @@ def test_get_sliding_window():
# For Qwen1.5/Qwen2, get_sliding_window() should be None
# when use_sliding_window is False.
qwen2_model_config = ModelConfig(
"Qwen/Qwen1.5-7B",
"Qwen/Qwen2.5-1.5B",
task="auto",
tokenizer="Qwen/Qwen1.5-7B",
tokenizer="Qwen/Qwen2.5-1.5B",
tokenizer_mode="auto",
trust_remote_code=False,
seed=0,
Expand Down
4 changes: 2 additions & 2 deletions tests/test_regression.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,8 +55,8 @@ def test_gc():


def test_model_from_modelscope(monkeypatch):
# model: https://modelscope.cn/models/qwen/Qwen1.5-0.5B-Chat/summary
MODELSCOPE_MODEL_NAME = "qwen/Qwen1.5-0.5B-Chat"
# model: https://modelscope.cn/models/qwen/Qwen2.5-1.5B-Instruct/summary
MODELSCOPE_MODEL_NAME = "qwen/Qwen2.5-1.5B-Instruct"
monkeypatch.setenv("VLLM_USE_MODELSCOPE", "True")
try:
llm = LLM(model=MODELSCOPE_MODEL_NAME)
Expand Down