frozenleaves/vllm
1
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2025-10-20 14:53:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

[CI Bugfix] Fix failing test_model_load_with_params tests due to tokenizer refactor (#25086)

Browse Source 
Signed-off-by: mgoin <mgoin64@gmail.com>
This commit is contained in:
Michael Goin
2025-09-17 14:15:05 -04:00
committed by GitHub
parent 9d442b7c48
commit e3db5ebb66
1 changed files with 5 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
tests/model_executor/test_model_load_with_params.py
View File

@ -47,8 +47,8 @@ def test_model_loading_with_params(vllm_runner, monkeypatch):
assert model_config.pooler_config.normalize
# asserts on the tokenizer loaded
assert model_tokenizer.tokenizer_id == "BAAI/bge-base-en-v1.5"
assert model_tokenizer.tokenizer.model_max_length == 512
assert model_config.tokenizer == "BAAI/bge-base-en-v1.5"
assert model_tokenizer.model_max_length == 512
def check_model(model):
assert isinstance(model, BertEmbeddingModel)
@ -87,8 +87,8 @@ def test_roberta_model_loading_with_params(vllm_runner, monkeypatch):
assert model_config.pooler_config.normalize
# asserts on the tokenizer loaded
assert model_tokenizer.tokenizer_id == "intfloat/multilingual-e5-base"
assert model_tokenizer.tokenizer.model_max_length == 512
assert model_config.tokenizer == "intfloat/multilingual-e5-base"
assert model_tokenizer.model_max_length == 512
def check_model(model):
assert isinstance(model, RobertaEmbeddingModel)
@ -116,8 +116,7 @@ def test_facebook_roberta_model_loading_with_params(vllm_runner, monkeypatch):
output = vllm_model.embed("Write a short story about a robot that"
" dreams for the first time.\n")
model_tokenizer = vllm_model.llm.llm_engine.tokenizer
assert model_tokenizer.tokenizer_id == model_name
assert vllm_model.llm.llm_engine.model_config.tokenizer == model_name
def check_model(model):
assert isinstance(model, RobertaEmbeddingModel)