Use official nvidia-ml-py instead of pynvml

9ffd8116f081b1f20600c95ab11b6b090819bf00 · 2024-10-05 14:46:19 +0200 · Gaetan Lepage

Files touched

M setup.py

Diff

commit 9ffd8116f081b1f20600c95ab11b6b090819bf00
Author: Gaetan Lepage <gaetan@glepage.com>
Date:   Sat Oct 5 14:46:19 2024 +0200

    Use official nvidia-ml-py instead of pynvml
---
 setup.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/setup.py b/setup.py
index 75d570e9..a8ad22bd 100644
--- a/setup.py
+++ b/setup.py
@@ -12,11 +12,11 @@ install_requires = [
   "Jinja2==3.1.4",
   "netifaces==0.11.0",
   "numpy==2.0.0",
+  "nvidia-ml-py==12.560.30",
   "pillow==10.4.0",
   "prometheus-client==0.20.0",
   "protobuf==5.27.1",
   "psutil==6.0.0",
-  "pynvml==11.5.3",
   "requests==2.32.3",
   "rich==13.7.1",
   "safetensors==0.4.3",

← 9b9f40d4 only stream results for the same request id. this allows mul · back to Exo · Point `llama-3.1-70b-bf16` model to the actually bf16 versio 82c7ce69 →