Handle new_block_ids is None.

[QiliangCui]

Description

torchax-jax run failed with error

�[1;36m(EngineCore_0 pid=278)�[0;0m ERROR 08-21 02:21:29 [core.py:710]     for block_ids, new_ids in zip(req_state.block_ids,
�[1;36m(EngineCore_0 pid=278)�[0;0m ERROR 08-21 02:21:29 [core.py:710]                               ^^^^^^^^^^^^^^^^^^^^^^^^
�[1;36m(EngineCore_0 pid=278)�[0;0m ERROR 08-21 02:21:29 [core.py:710] TypeError: 'NoneType' object is not iterable

This is because of PR: vllm-project/vllm#23262

The fix takes this change as reference.

If the change fixes a bug or a Github issue, please include a link, e.g.,:
FIXES: b/440212621

Tests

1. Manually run MODEL_IMPL_TYPE=vllm TPU_BACKEND_TYPE=jax python examples/offline_inference.py --model=meta-llama/Llama-3.1-8B-Instruct --tensor_parallel_size=1 --task=generate --max_model_len=1024 --download_dir /mnt/disks/persist
2. Wait for CIT: https://buildkite.com/tpu-commons/tpu-commons-ci/builds/2077

[github-actions]

Description

Start with a short description of what the PR does and how this is a change from
the past.

The rest of the description includes relevant details and context, examples:

• why is this change being made,
• the problem being solved and any relevant context,
• why this is a good solution,
• some information about the specific implementation,
• shortcomings of the solution and possible future improvements.

If the change fixes a bug or a Github issue, please include a link, e.g.,:
FIXES: b/123456
FIXES: #123456

Tests

Please describe how you tested this change, and include any instructions and/or
commands to reproduce.

Checklist

Before submitting this PR, please make sure:

• I have performed a self-review of my code.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have made or will make corresponding changes to any relevant documentation.