better readme + 3.14

Author: anakin87

.github/workflows/vllm.yml

@@ -32,8 +32,7 @@ env:
   VLLM_MODEL: "Qwen/Qwen3-0.6B"
   # we only test on Ubuntu to keep vLLM server running simple
   TEST_MATRIX_OS: '["ubuntu-latest"]'
-  # numba not compatible with Python 3.14
-  TEST_MATRIX_PYTHON: '["3.10", "3.13"]'
+  TEST_MATRIX_PYTHON: '["3.10", "3.14"]'
 
 jobs:
   compute-test-matrix:

integrations/vllm/README.md

@@ -10,3 +10,11 @@
 ## Contributing
 
 Refer to the general [Contribution Guidelines](https://github.com/deepset-ai/haystack-core-integrations/blob/main/CONTRIBUTING.md).
+
+To run integration tests locally, you need to have a running vLLM server. Refer to the [workflow file](https://github.com/deepset-ai/haystack-core-integrations/blob/main/.github/workflows/vllm.yml) for more details.
+
+For example, on macOs, you can install [vLLM-metal](https://github.com/vllm-project/vllm-metal) and run the server with:
+
+```bash
+source ~/.venv-vllm-metal/bin/activate && vllm serve Qwen/Qwen3-0.6B --reasoning-parser qwen3 --max-model-len 1024 --enforce-eager --dtype bfloat16 --enable-auto-tool-choice --tool-call-parser hermes
+```