[ci] add pytorch kvint testcase into function regresstion

[zhulinJulia24]

after #2438

1. [tested] add pytorch kvint testcase
2. [tested] review test config according to support models, most models supported by turbomind are compatible with kvint and 4-bit quantization. Modify the configuration YAML file to make it easy to maintain.
3. [testing] add llama3.2 model into pytorch backend

[zhulinJulia24]

@lvhan028 Qwen2-VL does not supported by turbomind backend

error is

lmdeploy convert qwen /nvme/qa_test_models/Qwen/Qwen2-VL-2B-Instruct --dst-path /nvme/qa_test_models/autotest_model/workspace_Qwen/Qwen2-VL-2B-Instruct

2024-10-12 09:44:10,786 - lmdeploy - WARNING - converter.py:324 - The argument `<model_name>` is deprecated and unused now. It will be removed on 2024.12.31. It was originally used to specify the name of the built-in chat template, but now it is substituted with a clearer parameter `--chat-template`
Unrecognized keys in `rope_scaling` for 'rope_type'='default': {'mrope_section'}
Traceback (most recent call last):
  File "/opt/py3/bin/lmdeploy", line 8, in <module>
    sys.exit(run())
  File "/opt/py3/lib/python3.10/site-packages/lmdeploy/cli/entrypoint.py", line 42, in run
    args.run(args)
  File "/opt/py3/lib/python3.10/site-packages/lmdeploy/cli/cli.py", line 165, in convert
    main(**kwargs)
  File "/opt/py3/lib/python3.10/site-packages/lmdeploy/turbomind/deploy/converter.py", line 334, in main
    assert is_supported(model_path), (
AssertionError: turbomind does not support /nvme/qa_test_models/Qwen/Qwen2-VL-2B-Instruct. Plz try pytorch engine instead.

Should we update this table