[CI/Build] bump minimum cmake version

[dtrifiro]

• find_python_from_executable uses find_package(PYTHON ... Development.SABIModule), which is only available since cmake>=3.26
• cmake and ninja are build-time requirements, so they should not be in requirements-common.txt or requirements-openvino.txt

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which consists a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of default ones by unblocking the steps in your fast-check build on Buildkite UI.

Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge).

To run full CI, you can do one of these:

• Comment /ready on the PR
• Add ready label to the PR
• Enable auto-merge.

🚀

[dtrifiro]

Some extra context for this: having an older cmake version would result in a cryptic error when building

  Unable to find python matching:
  /home/dtrifiro/work/vllm/.venv/bin/python.
Call Stack (most recent call first):
  CMakeLists.txt:43 (find_python_from_executable)

[bnellnm]

LGTM!

[dtrifiro]

CI failures seem unrelated:

test_traces
...
ValueError: OpenTelemetry packages must be installed before configuring 'otlp_traces_endpoint'

[dtrifiro]

/ready

[dtrifiro]

Anything holding this back?

[mgoin]

@dtrifiro looking at the failing check logs, it seems the Intel CPU test cannot find its cmake executable with this change. It would be great if you could investigate this

[dtrifiro]

@mgoin I opened #7298 to take care of the Dockerfile.cpu issues, this should go through once that's merged.

[dtrifiro]

Looking at the failure logs, it seems that they're unrelated:

entrypoints/openai/test_completion.py::test_single_completion[-zephyr-pa-8] INFO 08-09 08:48:57 logger.py:36] Received request cmpl-c29cba85cf3a4c7a930f4aa562d54614-0: prompt: 'Hello, my name is', params: SamplingParams(n=1, best_of=1, presence_penalty=0.0, frequency_penalty=0.0, repetition_penalty=1.0, temperature=0.0, top_p=1.0, top_k=-1, min_p=0.0, seed=None, use_beam_search=False, length_penalty=1.0, early_stopping=False, stop=[], stop_token_ids=[], include_stop_str_in_output=False, ignore_eos=False, max_tokens=5, min_tokens=0, logprobs=None, prompt_logprobs=None, skip_special_tokens=True, spaces_between_special_tokens=True, truncate_prompt_tokens=None), prompt_token_ids: [1, 22557, 28725, 586, 1141, 349], lora_request: None, prompt_adapter_request: PromptAdapterRequest(prompt_adapter_name='zephyr-pa', prompt_adapter_id=1, prompt_adapter_local_path='/root/.cache/huggingface/hub/models--swapnilbp--llama_tweet_ptune/snapshots/ad7faa5e642dd3f4311c927b0c423402444cd7a8', prompt_adapter_num_virtual_tokens=8).
[2024-08-09T08:48:57Z] INFO 08-09 08:48:57 async_llm_engine.py:199] Added request cmpl-c29cba85cf3a4c7a930f4aa562d54614-0.
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61] Engine background task failed
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61] Traceback (most recent call last):
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 51, in _log_task_completion
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     return_value = task.result()
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 772, in run_engine_loop
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     result = task.result()
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 715, in engine_step
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     request_outputs = await self.engine.step_async(virtual_engine)
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 282, in step_async
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     output = await self.model_executor.execute_model_async(
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/executor/gpu_executor.py", line 159, in execute_model_async
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     output = await make_async(self.driver_worker.execute_model
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/lib/python3.10/concurrent/futures/thread.py", line 58, in run
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     result = self.fn(*self.args, **self.kwargs)
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/worker_base.py", line 236, in execute_model
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     self.model_runner.prepare_model_input(
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1435, in prepare_model_input
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     model_input = self._prepare_model_input_tensors(
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1006, in _prepare_model_input_tensors
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     builder.add_seq_group(seq_group_metadata)
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 638, in add_seq_group
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     per_seq_group_fn(inter_data, seq_group_metadata)
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 590, in _compute_prompt_adapter_input
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61]     assert inter_data.n_seqs == 1
[2024-08-09T08:48:57Z] ERROR 08-09 08:48:57 async_llm_engine.py:61] AssertionError
[2024-08-09T08:48:57Z] Exception in callback _log_task_completion(error_callback=<bound method...7fec46856bc0>>)(<Task finishe...ertionError()>) at /usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py:41
[2024-08-09T08:48:57Z] handle: <Handle _log_task_completion(error_callback=<bound method...7fec46856bc0>>)(<Task finishe...ertionError()>) at /usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py:41>
[2024-08-09T08:48:57Z] Traceback (most recent call last):
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 51, in _log_task_completion
[2024-08-09T08:48:57Z]     return_value = task.result()
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 772, in run_engine_loop
[2024-08-09T08:48:57Z]     result = task.result()
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 715, in engine_step
[2024-08-09T08:48:57Z]     request_outputs = await self.engine.step_async(virtual_engine)
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 282, in step_async
[2024-08-09T08:48:57Z]     output = await self.model_executor.execute_model_async(
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/executor/gpu_executor.py", line 159, in execute_model_async
[2024-08-09T08:48:57Z]     output = await make_async(self.driver_worker.execute_model
[2024-08-09T08:48:57Z]   File "/usr/lib/python3.10/concurrent/futures/thread.py", line 58, in run
[2024-08-09T08:48:57Z]     result = self.fn(*self.args, **self.kwargs)
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/worker_base.py", line 236, in execute_model
[2024-08-09T08:48:57Z]     self.model_runner.prepare_model_input(
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1435, in prepare_model_input
[2024-08-09T08:48:57Z]     model_input = self._prepare_model_input_tensors(
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1006, in _prepare_model_input_tensors
[2024-08-09T08:48:57Z]     builder.add_seq_group(seq_group_metadata)
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 638, in add_seq_group
[2024-08-09T08:48:57Z]     per_seq_group_fn(inter_data, seq_group_metadata)
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 590, in _compute_prompt_adapter_input
[2024-08-09T08:48:57Z]     assert inter_data.n_seqs == 1
[2024-08-09T08:48:57Z] AssertionError
[2024-08-09T08:48:57Z]
[2024-08-09T08:48:57Z] The above exception was the direct cause of the following exception:
[2024-08-09T08:48:57Z]
[2024-08-09T08:48:57Z] Traceback (most recent call last):
[2024-08-09T08:48:57Z]   File "/usr/lib/python3.10/asyncio/events.py", line 80, in _run
[2024-08-09T08:48:57Z]     self._context.run(self._callback, *self._args)
[2024-08-09T08:48:57Z]   File "/usr/local/lib/python3.10/dist-packages/vllm/engine/async_llm_engine.py", line 63, in _log_task_completion
[2024-08-09T08:48:57Z]     raise AsyncEngineDeadError(
[2024-08-09T08:48:57Z] vllm.engine.async_llm_engine.AsyncEngineDeadError: Task finished unexpectedly. This should never happen! Please open an issue on Github. See stack trace above for the actual cause.
[2024-08-09T08:48:57Z] WARNING 08-09 08:48:57 server.py:103] Failed to abort request cmpl-c29cba85cf3a4c7a930f4aa562d54614-0

[njhill]

Thanks @dtrifiro