Main features:

1. Use the ComputingResourceManager to manage all computing workers.
2. Ensure fair worker number between summary directories at first. So every summary directories in cache will be loaded simultaneously.
3. When a summary directory is loaded, it's worker will be released, and other unfinished summary directories can use the released workers to speed up. This way we solved the slow worker problem.

Code changes:

1. Added computing_resource_mgr.py
2. Passed ComputingResourceManager instances instead of workers_count
3. Simplified the _load_single_fine() function a bit.

- collect_cpu
- collect_mem
- collect_npu

Support tensor visualization. 1.Tensor display in a table, it can support no more than two dimensions tensor visualization; 2.Tensor histogram visualization for all step in cache.

This reverts commit a828f668.

Modify points: 1. adapt the task id change; 2. support parsing point
files; 3. adapt the source path change

Collect timeline information after training finished.
Support timeline display on web UI.

Use the AST replaces the importlib/inspect modules to analyze and modify network definition script.
The importlib/inspect must load python script to analyze, but AST analysis is static code analysis and is very secure.

add op mappings

fix pylint

fix mistake { as [

add tests

add test for eye and revise func and mappings

revise mapping and add test

fix pylint

fix pylint

fix the bug that when the profiler parameter subgraph is Default or Gradients, the profiler analyse will raise an exception