[Good First Issue][NNCF]: Replace common tensor_statistics by experimental tensor_statistics

[kshpv]

Context

Historically, NNCF has maintained two sets of tensor statistics: the old tensor statistics and the new tensor statistics. The new tensor statistics, located in the experimental/common/tensor_statistics directory, offer improved functionality and better performance. However, the codebase still contains references to the old tensor statistics, leading to redundancy and potential confusion.

What needs to be done?

1. Search the codebase for all instances where the old tensor statistics are being used. This includes algorithms, tests, and any utility functions.
2. Modify the identified algorithms to utilize the new tensor statistics from experimental/common/tensor_statistics.
3. Update the corresponding tests to ensure they are compatible with the new tensor statistics.
4. Replace the contents of common/tensor_statistics/collectors with the corresponding implementations from experimental/common/tensor_statistics.
5. Once all references to the old tensor statistics have been updated, remove the experimental/common/tensor_statistics.
6. Remove common/tensor_statistics/reduction if it is not needed anymore.

In the end, there should be no experimental/common/tensor_statistics.

Example Pull Requests

#2117

Resources

• Contribution guide - start here!
• Intel DevHub Discord channel - engage in discussions, ask questions and talk to OpenVINO developers
• How to link your Pull Request to an issue

Contact points

@kshpv

[olegkkruglov]

.take

[github-actions]

Thank you for looking into this issue! Please let us know if you have any questions or require any help.