[CI] chore: Update e2e_ascend CI config

[FightingZhen]

What does this PR do?

This PR is committed for solving 3 things:

1. All test cases in e2e_ascend CI pipeline use 8 NPUs by default, which prevents the machine's original performance from being fully utilized. This PR is committed for solving this problem. Thank @zheliuyu for finding this problem :)
2. Remove qwen3 grpo test case in e2e_ascend.yml because it is similar to qwen2.5 grpo.
3. Remove ppo test case in e2e_ascend.yml because it is not work since first commit [model, ci] feat: add qwen3-8b ppo script on ASCEND NPU #3502 , @xvxuopop is working for solving this.
4. Update e2e_ascend CI scan path for covering most of file modification case.
5. Ignore loading libnuma.so for Ascend NPU.
6. Fix qwen2_vl flash-attention related functions unavailable error for Ascend NPU.

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, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data
  • 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

Not related.

API and Usage Example

Not related.

Design & Code Changes

Not related.

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).)

[gemini-code-assist]

Code Review

This pull request updates the number of GPUs/NPUs from 8 to 16 across several test scripts to better utilize hardware resources. The changes are correct, but I've identified an opportunity to improve maintainability and consistency. In several scripts, the number of GPUs is hardcoded. I've suggested using an environment variable with a default value, which makes the scripts more flexible and aligns with the pattern already present in one of the modified files.