Skip to content

GitLab

Switch branch/tag
Find file Normal viewHistoryPermalink
requirements-rocm.txt 366 Bytes
Newer Older
bnellnm's avatar
Cmake based build system (#2830)  
bnellnm committed
1 
cmake>=3.21
TJian's avatar
Merge EmbeddedLLM/vllm-rocm into vLLM main (#1836)  
TJian committed
2
3
4 
ninja  # For faster builds.
typing-extensions>=4.8.0
starlette
yhu422's avatar
Usage Stats Collection (#2852)  
yhu422 committed
5
6 
requests
py-cpuinfo
TJian's avatar
Merge EmbeddedLLM/vllm-rocm into vLLM main (#1836)  
TJian committed
7 
psutil
Hongxia Yang's avatar
[ROCm][Bugfix] Fixed several bugs related to rccl path and attention selector logic (#3699)  
Hongxia Yang committed
8 
ray == 2.9.3
TJian's avatar
Merge EmbeddedLLM/vllm-rocm into vLLM main (#1836)  
TJian committed
9
10
11 
sentencepiece  # Required for LLaMA tokenizer.
numpy
tokenizers>=0.15.0
Roger Wang's avatar
[Misc] Bump transformers version (#3592)  
Roger Wang committed
12 
transformers >= 4.39.1  # Required for StarCoder2 & Llava.
TJian's avatar
Merge EmbeddedLLM/vllm-rocm into vLLM main (#1836)  
TJian committed
13
14 
fastapi
uvicorn[standard]
Jannis Schönleber's avatar
migrate pydantic from v1 to v2 (#2531)  
Jannis Schönleber committed
15 
pydantic >= 2.0  # Required for OpenAI server.
Allen.Dou's avatar
Restrict prometheus_client >= 0.18.0 to prevent errors when importing pkgs (#3070)  
Allen.Dou committed
16 
prometheus_client >= 0.18.0
Roger Wang's avatar
[Misc] Bump transformers version (#3592)  
Roger Wang committed
17 
outlines == 0.0.34