[ci] chore: npu ci use cann9.0.0

[daikang6]

What does this PR do?

update npu's cann to 9.0.0 version

Checklist Before Starting

• Search for similar PRs. Paste at least one query link here: ...
• Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
  • {modules} include fsdp, megatron, veomni, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data, cfg, reward, fully_async, one_step_off
  • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
  • {type} is in feat, fix, refactor, chore, test
  • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
  • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

Design & Code Changes

Demonstrate the high-level design if this PR is complex, and list the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

• Read the Contribute Guide.
• Apply pre-commit checks: pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always
• Add / Update the documentation.
• Add unit or end-to-end test(s) to the CI workflow to cover all the code. If not feasible, explain why: ...
• Once your PR is ready for CI, send a message in the ci-request channel in the verl Slack workspace. (If not accessible, please try the Feishu group (飞书群).)
• If your PR is related to the recipe submodule, please also update the reference to the submodule commit via git submodule update --remote or cd recipe && git pull origin main.

[gemini-code-assist]

Code Review

This pull request updates the Ascend CI documentation to use a newer vLLM image version, configures the GRPO Qwen3-8B test script to calculate log probabilities, and simplifies the output log path for the PPO Qwen3-8B test script. A review comment points out that the log directory for the PPO script might not exist, which could cause the tee command to fail and result in lost CI logs. It is recommended to ensure the directory is created prior to writing the log.

[gemini-code-assist]

If the directory /root/.cache/nightly_log/ppo_qwen3_8b does not exist, the tee command will fail to write the log file, resulting in lost CI logs. To prevent this, ensure the directory is created before writing the log. Additionally, redirect stderr to stdout using 2>&1 before piping to tee to ensure that error messages are captured in the log file.

Suggested change

	trainer.total_training_steps=15 2>&1 | tee /root/.cache/nightly_log/ppo_qwen3_8b/ppo_qwen3-8b_fsdp_npu.log
	trainer.total_training_steps=15 2>&1 | { mkdir -p /root/.cache/nightly_log/ppo_qwen3_8b && tee /root/.cache/nightly_log/ppo_qwen3_8b/ppo_qwen3-8b_fsdp_npu.log; }

References

1. Ensure that the log directory exists before writing to it in shell scripts.
2. In shell scripts, redirect stderr to stdout using 2>&1 before piping to tee to ensure that error messages are captured in the log file.

[daikang6]

npu unit test加上
import verl.utils.device as device_module
monkeypatch.setattr(device_module, "is_npu_available", False)