diff --git a/computility-run.yaml b/computility-run.yaml index 0be4999a..02adf99e 100644 --- a/computility-run.yaml +++ b/computility-run.yaml @@ -51,3 +51,7 @@ env: value: /tmp/vllm-request-metrics.jsonl - name: VLLM_CACHE_BLOCK_SIZE value: '16' + - name: PYTORCH_CUDA_ALLOC_CONF + value: max_split_size_mb:512 + - name: OMP_NUM_THREADS + value: '1'