biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix][v1] Fix step pooler implementation and step pooling usage in v1 (#19956)

Browse Source 
Signed-off-by: Isotr0py <2037008807@qq.com>
This commit is contained in:
Isotr0py
2025-06-24 02:38:06 +08:00
committed by GitHub
parent 68aaeb3749
commit 61f4fc5dc6
14 changed files with 164 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tests/model_executor/test_model_load_with_params.py
View File

@@ -29,8 +29,8 @@ def test_model_loading_with_params(vllm_runner):
revision=REVISION,
dtype="float16",
max_model_len=MAX_MODEL_LEN) as vllm_model:
output = vllm_model.encode("Write a short story about a robot that"
" dreams for the first time.\n")
output = vllm_model.embed("Write a short story about a robot that"
" dreams for the first time.\n")
model_config = vllm_model.model.llm_engine.model_config
model_tokenizer = vllm_model.model.llm_engine.tokenizer
@@ -67,8 +67,8 @@ def test_roberta_model_loading_with_params(vllm_runner):
revision=REVISION_ROBERTA,
dtype="float16",
max_model_len=MAX_MODEL_LEN) as vllm_model:
output = vllm_model.encode("Write a short story about a robot that"
" dreams for the first time.\n")
output = vllm_model.embed("Write a short story about a robot that"
" dreams for the first time.\n")
model_config = vllm_model.model.llm_engine.model_config
model_tokenizer = vllm_model.model.llm_engine.tokenizer
@@ -105,8 +105,8 @@ def test_facebook_roberta_model_loading_with_params(vllm_runner):
with vllm_runner(model_name=model_name,
dtype="float16",
max_model_len=MAX_MODEL_LEN) as vllm_model:
output = vllm_model.encode("Write a short story about a robot that"
" dreams for the first time.\n")
output = vllm_model.embed("Write a short story about a robot that"
" dreams for the first time.\n")
model_tokenizer = vllm_model.model.llm_engine.tokenizer
assert model_tokenizer.tokenizer_id == model_name