[codex] Fix shared embeddings and string sparse validation

[shenweichen]

Summary

• Reuse embedding layers by embedding_name for both SparseFeat and VarLenSparseFeat instead of recreating and overwriting duplicate entries.
• Preserve sequence masking when a shared embedding is used by varlen sparse features.
• Raise a clear ValueError when a string SparseFeat is configured without use_hash=True, so users do not hit TensorFlow's lower-level string-to-int embedding lookup failure.
• Add regression coverage for shared embeddings and ESMM string sparse feature configuration.

Why

This addresses #419 by making shared embedding semantics explicit and avoiding repeated initialization for the same embedding_name. It also addresses #548 by clarifying the supported path for string sparse features in ESMM and other models: use use_hash=True, or pre-encode values to integer ids before passing them to DeepCTR.

Validation

• git diff --check
• python -m compileall -q deepctr tests
• python -m pytest -q tests/feature_test.py tests/models/MTL_test.py was attempted locally, but this environment does not have TensorFlow installed (ModuleNotFoundError: No module named 'tensorflow').

[codecov]

Codecov Report

❌ Patch coverage is 93.75000% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 78.23%. Comparing base (b6a623a) to head (28035a0).
⚠️ Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
deepctr/feature_column.py	77.77%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #551      +/-   ##
==========================================
+ Coverage   78.11%   78.23%   +0.11%     
==========================================
  Files          61       61              
  Lines        3304     3331      +27     
==========================================
+ Hits         2581     2606      +25     
- Misses        723      725       +2     

Flag	Coverage Δ
pytest	78.23% <93.75%> (+0.11%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ 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.