Set max-model-len to 32768 for testing

2026-08-09 22:19:32 -07:00
parent 6bdabfc65b
commit 11d66c3823
1 changed files with 1 additions and 1 deletions
@@ -25,7 +25,7 @@ exec docker run --rm --name vllm-awq-qwen36-dflash-ampere \
    vllm/vllm-openai:latest "${MODEL}" \
    --served-model-name "frank-default-small" \
    --max-num-seqs 1 \
-    --max-model-len 212992 \
+    --max-model-len 32768 \
    --tensor-parallel-size 1 \
    --enable-chunked-prefill \
    --max-num-batched-tokens 4096 \