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

[Model] Pooling models default to using chunked prefill & prefix caching if supported. (#20930)

Browse Source 
Signed-off-by: wang.yuqi <noooop@126.com>
This commit is contained in:
wang.yuqi
2025-08-12 00:41:37 +08:00
committed by GitHub
parent 16fb668b61
commit 84cf78acee
31 changed files with 452 additions and 261 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
tests/models/language/pooling/test_nomic.py
View File

@@ -3,22 +3,23 @@
import pytest
from .embed_utils import EmbedModelInfo, correctness_test_embed_models
from ...utils import CLSPoolingEmbedModelInfo, EmbedModelInfo
from .embed_utils import correctness_test_embed_models
from .mteb_utils import mteb_test_embed_models
MODELS = [
EmbedModelInfo("nomic-ai/nomic-embed-text-v1",
architecture="NomicBertModel",
enable_test=True),
EmbedModelInfo("nomic-ai/nomic-embed-text-v1.5",
architecture="NomicBertModel",
enable_test=False),
EmbedModelInfo("nomic-ai/CodeRankEmbed",
architecture="NomicBertModel",
enable_test=False),
EmbedModelInfo("nomic-ai/nomic-embed-text-v2-moe",
architecture="NomicBertModel",
enable_test=True)
CLSPoolingEmbedModelInfo("nomic-ai/nomic-embed-text-v1",
architecture="NomicBertModel",
enable_test=True),
CLSPoolingEmbedModelInfo("nomic-ai/nomic-embed-text-v1.5",
architecture="NomicBertModel",
enable_test=False),
CLSPoolingEmbedModelInfo("nomic-ai/CodeRankEmbed",
architecture="NomicBertModel",
enable_test=False),
CLSPoolingEmbedModelInfo("nomic-ai/nomic-embed-text-v2-moe",
architecture="NomicBertModel",
enable_test=True)
]