Support reproducible ordering in Petastorm

[arushi297]

This PR addresses the issue where the reader doesn't give reproducible data order for multiple workers when shuffling is disabled or seed is set. The following changes are made to address the issue:

1. Ventilator Queue Modifications: Each worker now has a separate ventilator queue, ensuring deterministic data ordering through a round-robin strategy and eliminating race conditions.

2. np.default_rng(): The switch from np.random.RandomState() to np.default_rng() provides reproducible shuffling, ensuring consistent outcomes when a seed is provided.

3. Result Queue Improvements: Individual result queues for each worker have been implemented. Results are collected in a round-robin manner.

The following behavior is expected as a result of this change:

1. Shuffling = False, Deterministically natural order
2. Shuffling = True, Seed = Provided, Deterministically randomized order
3. Shuffling = True, Seed = Not Provided (or 0), Non-deterministically randomized order

Testing and Validation: Added unit tests to cover the 3 scenario stated. Also, performed experiments that confirmed that data order remains consistent when shuffling is disabled and is reproducible with a seed, improving the reliability of dataloader.

[codecov]

Codecov Report

❌ Patch coverage is 94.64286% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 86.22%. Comparing base (5f2797c) to head (70d42ea).
⚠️ Report is 2 commits behind head on master.

Files with missing lines	Patch %	Lines
petastorm/workers_pool/thread_pool.py	91.17%	2 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #810      +/-   ##
==========================================
- Coverage   86.26%   86.22%   -0.05%     
==========================================
  Files          84       84              
  Lines        5090     5109      +19     
  Branches      790      801      +11     
==========================================
+ Hits         4391     4405      +14     
- Misses        560      562       +2     
- Partials      139      142       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.