Allow generation of ParamsWithStats from FastLDF plus parameters

[penelopeysm]

This has no real impact on DynamicPPL, but using this in Turing should lead to some nice speedups because this:

1. is faster to evaluate than regular unflatten + evaluate!!;
2. avoids deepcopying VarInfo.

Closes #1119.

[github-actions]

Benchmark Report for Commit c475f96

Computer Information

Julia Version 1.11.7
Commit f2b3dbda30a (2025-09-08 12:10 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 7763 64-Core Processor
  WORD_SIZE: 64
  LLVM: libLLVM-16.0.6 (ORCJIT, znver3)
Threads: 1 default, 0 interactive, 1 GC (on 4 virtual cores)

Benchmark Results

┌───────────────────────┬───────┬─────────────┬───────────────────┬────────┬────────────────┬─────────────────┐
│                 Model │   Dim │  AD Backend │           VarInfo │ Linked │ t(eval)/t(ref) │ t(grad)/t(eval) │
├───────────────────────┼───────┼─────────────┼───────────────────┼────────┼────────────────┼─────────────────┤
│ Simple assume observe │     1 │ forwarddiff │             typed │  false │            5.8 │             1.8 │
│           Smorgasbord │   201 │ forwarddiff │             typed │  false │          632.4 │            43.6 │
│           Smorgasbord │   201 │ forwarddiff │ simple_namedtuple │   true │          368.7 │            53.9 │
│           Smorgasbord │   201 │ forwarddiff │           untyped │   true │          706.2 │            35.4 │
│           Smorgasbord │   201 │ forwarddiff │       simple_dict │   true │         6210.6 │            24.3 │
│           Smorgasbord │   201 │ forwarddiff │      typed_vector │   true │          670.4 │            41.2 │
│           Smorgasbord │   201 │ forwarddiff │    untyped_vector │   true │          697.2 │            40.9 │
│           Smorgasbord │   201 │ reversediff │             typed │   true │          795.9 │            45.5 │
│           Smorgasbord │   201 │    mooncake │             typed │   true │          637.8 │             6.0 │
│           Smorgasbord │   201 │      enzyme │             typed │   true │          797.8 │             3.8 │
│    Loop univariate 1k │  1000 │    mooncake │             typed │   true │         3443.9 │             5.6 │
│       Multivariate 1k │  1000 │    mooncake │             typed │   true │          879.5 │             8.9 │
│   Loop univariate 10k │ 10000 │    mooncake │             typed │   true │        39292.5 │             5.2 │
│      Multivariate 10k │ 10000 │    mooncake │             typed │   true │         7654.6 │             9.8 │
│               Dynamic │    10 │    mooncake │             typed │   true │          106.0 │            11.7 │
│              Submodel │     1 │    mooncake │             typed │   true │            7.6 │             6.6 │
│                   LDA │    12 │ reversediff │             typed │   true │          886.4 │             2.0 │
└───────────────────────┴───────┴─────────────┴───────────────────┴────────┴────────────────┴─────────────────┘

[github-actions]

DynamicPPL.jl documentation for PR #1129 is available at:
https://TuringLang.github.io/DynamicPPL.jl/previews/PR1129/

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.94%. Comparing base (4ec0c72) to head (c475f96).

Additional details and impacted files

@@              Coverage Diff               @@
##           py/fastldf    #1129      +/-   ##
==============================================
+ Coverage       81.86%   81.94%   +0.08%     
==============================================
  Files              41       41              
  Lines            3949     3961      +12     
==============================================
+ Hits             3233     3246      +13     
+ Misses            716      715       -1     

☔ 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.