ASCEND TORCH_NPU INFERENCE performance improvement optimization point

[lightnateriver]

FOR ASCEND TORCH_NPU BACKEND: By the following configuration, private conv format is not allowed, which reduces format conversion and optimizes the speed of the conv operator. It can also avoid the possible functional issues caused by the propagation of private formats and the potential increase in memory usage. The verification has shown a performance improvement, so it can be considered for inclusion in the main branch：

`if name == "main":

torch.npu.config.allow_internal_format = False`

[lightnateriver]

@patil-suraj @kashif

[DN6]

@lightnateriver sorry what is the change you're proposing here exactly?

[lightnateriver]

这是一封自动回复邮件。已经收到您的来信，我会尽快回复。白茶清欢无别事,我在等风也等你。清酒独酌了无趣,我在梦花也梦你。

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[lightnateriver]

这是一封自动回复邮件。已经收到您的来信，我会尽快回复。白茶清欢无别事,我在等风也等你。清酒独酌了无趣,我在梦花也梦你。

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[lightnateriver]

这是一封自动回复邮件。已经收到您的来信，我会尽快回复。白茶清欢无别事,我在等风也等你。清酒独酌了无趣,我在梦花也梦你。

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[lightnateriver]

这是一封自动回复邮件。已经收到您的来信，我会尽快回复。白茶清欢无别事,我在等风也等你。清酒独酌了无趣,我在梦花也梦你。