A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 17, 04:15 UTC
[Misc] Bump transformers version (#3592)
Roger Wang committed
Mar 24, 2024 at 06:32 UTC
f8a12ecc7f7ebcffe26e1ae405c6aa533fc400cd
2 files changed
+3
-3
requirements-rocm.txt
+2
-2
@@ -7,9 +7,9 @@ ray >= 2.9
7
sentencepiece # Required for LLaMA tokenizer.
8
numpy
9
tokenizers>=0.15.0
10
-transformers >= 4.39.0 # Required for StarCoder2.
10
+transformers >= 4.39.1 # Required for StarCoder2 & Llava.
11
fastapi
12
uvicorn[standard]
13
pydantic >= 2.0 # Required for OpenAI server.
14
prometheus_client >= 0.18.0
15
-outlines == 0.0.34
\ No newline at end of file
15
+outlines == 0.0.34
requirements.txt
+1
-1
@@ -5,7 +5,7 @@ ray >= 2.9
5
sentencepiece # Required for LLaMA tokenizer.
6
numpy
7
torch == 2.1.2
8
-transformers >= 4.39.0 # Required for StarCoder2.
8
+transformers >= 4.39.1 # Required for StarCoder2 & Llava.
9
xformers == 0.0.23.post1 # Required for CUDA 12.1.
10
fastapi
11
uvicorn[standard]