[CI][Benchmark] Optimize performance benchmark workflow (vllm-project#1039)

This is a post patch of vllm-project#1014, for some convenience optimization
- Set cached dataset path for speed
- Use pypi to install escli-tool
- Add benchmark results convert script to have a developer-friendly
result
- Patch the `benchmark_dataset.py` to disable streaming load for
internet
- Add more trigger ways for different purpose, `pr` for debug,
`schedule` for daily test, `dispatch` and `pr-labled` for manual testing
of a single(current) commit
- Disable latency test for `qwen-2.5-vl`, (This script does not support
multi-modal yet)

No

CI passed

---------

Signed-off-by: wangli <[email protected]>
Signed-off-by: wangxiaoxin (A) <[email protected]>

Author: Potabk

.github/workflows/nightly_benchmarks.yaml

@@ -89,8 +89,11 @@ jobs:
 
       - name: Checkout vllm-project/vllm-ascend repo
         uses: actions/checkout@v4
+<<<<<<< HEAD
         with:
           fetch-depth: 0
+=======
+>>>>>>> fd6d021 ([CI][Benchmark] Optimize performance benchmark workflow (#1039))
 
       - name: Checkout vllm-project/vllm repo
         uses: actions/checkout@v4
@@ -136,7 +139,11 @@ jobs:
       - name: Install elastic_tool
         if: github.event_name == 'schedule'
         run: |
+<<<<<<< HEAD
           pip install escli-tool==0.2.1
+=======
+          pip install escli-tool==0.2.0
+>>>>>>> fd6d021 ([CI][Benchmark] Optimize performance benchmark workflow (#1039))
 
       - name: Collect pr info from vllm-project/vllm-ascend
         if: github.event_name == 'schedule'