fix: enable logprobs during spec decoding by default

Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>

Dockerfile.ubi

@@ -183,6 +183,11 @@ RUN --mount=type=cache,target=/root/.cache/pip \
 ENV HF_HUB_OFFLINE=1 \
     PORT=8000 \
     HOME=/home/vllm \
+    # As an optmization, vLLM disables logprobs when using spec decoding by
+    # default, but this would be unexpected to users of a hosted model that
+    # happens to have spec decoding
+    # see: https://github.com/vllm-project/vllm/pull/6485
+    DISABLE_LOGPROBS_DURING_SPEC_DECODING=false \
     # Allow requested max length to exceed what is extracted from the
     # config.json
     # see: https://github.com/vllm-project/vllm/pull/7080