stephane.david/vllm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True env

Browse Source
This commit is contained in:
thedarty
2026-06-19 16:26:39 +02:00
parent ebbe43deb6
commit 461690f0c1
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docker-compose.yml
+2 -2
View File
@@ -5,7 +5,7 @@ services:
network_mode: bridge network_mode: bridge
environment: environment:
- HF_TOKEN=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 - HF_TOKEN=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
- PYTORCH_ALLOC_CONF=expandable_segments:True,throw_on_cudamalloc_oom:True - PYTORCH_ALLOC_CONF=expandable_segments:True
ipc: host ipc: host
ports: ports:
- "8000:8000" - "8000:8000"
@@ -19,4 +19,4 @@ services:
count: all count: all
capabilities: [ gpu ] capabilities: [ gpu ]
runtime: nvidia runtime: nvidia
command: --model NVFP4/Qwen3-Coder-30B-A3B-Instruct-FP4 # command: --model NVFP4/Qwen3-Coder-30B-A3B-Instruct-FP4