Skip to content

Commit 3a37457

Browse files
committed
升级 lmdeploy==0.9.1 vllm==0.9.2 sglang[all]>=0.4.9.post3 transformers==4.53.3 xgrammar==0.1.21
1 parent 4f2de19 commit 3a37457

File tree

4 files changed

+313
-430
lines changed

4 files changed

+313
-430
lines changed

gpt_server/model_worker/embedding_infinity.py

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -48,9 +48,10 @@ def __init__(
4848
device = "cuda"
4949
logger.warning(f"使用{device}加载...")
5050
model_type = getattr(self.model_config, "model_type", None)
51-
bettertransformer = True
52-
if model_type is not None and "deberta" in model_type:
53-
bettertransformer = False
51+
bettertransformer = False
52+
# TODO bettertransformer = True transformer 出问题
53+
# if model_type is not None and "deberta" in model_type:
54+
# bettertransformer = False
5455
engine_args = EngineArgs(
5556
model_name_or_path=model_path,
5657
engine="torch",

pyproject.toml

Lines changed: 6 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
[project]
22
name = "gpt_server"
3-
version = "0.5.3"
3+
version = "0.5.4"
44
description = "gpt_server是一个用于生产级部署LLMs或Embedding的开源框架。"
55
readme = "README.md"
66
license = { text = "Apache 2.0" }
@@ -12,20 +12,20 @@ dependencies = [
1212
"ffmpy",
1313
"fschat==0.2.36",
1414
"infinity-emb[all]==0.0.76",
15-
"lmdeploy==0.9.0",
15+
"lmdeploy==0.9.1",
1616
"loguru>=0.7.2",
1717
"openai==1.86.0",
1818
"setuptools==75.2.0",
1919
"streamlit==1.39.0",
2020
"torch==2.6.0",
2121
"torchvision==0.20.1",
22-
"vllm==0.9.1",
22+
"vllm==0.9.2",
2323
"qwen_vl_utils",
2424
"evalscope[perf,rag]==0.16.1",
2525
"modelscope==1.26.0",
2626
"edge-tts>=7.0.0",
2727
"funasr>=1.2.6",
28-
"sglang[all]>=0.4.8.post1",
28+
"sglang[all]>=0.4.9.post3",
2929
"flashinfer-python",
3030
"flashtts>=0.1.7",
3131
"diffusers>=0.33.1",
@@ -39,10 +39,9 @@ override-dependencies = [
3939
"torchaudio==2.7.1",
4040
"torch==2.7.0",
4141
"triton",
42-
"outlines==0.1.11",
43-
"transformers==4.51.3", # infinity-emb
42+
"transformers==4.53.3", # infinity-emb
4443
"soundfile==0.13.1", # infinity
45-
"xgrammar==0.1.19", # sglang[all]==0.4.5 depends on xgrammar==0.1.17
44+
"xgrammar==0.1.21", # sglang[all]==0.4.5 depends on xgrammar==0.1.17
4645
]
4746

4847
[project.scripts]

0 commit comments

Comments
 (0)