26122874472
72%
main: 72%

Ran 19 May 2026 08:23PM UTC

Jobs 1

Files 31

Run time 1min

Badge

Embed ▾

Committed 19 May 2026 08:19PM UTC coverage: 72.219%. Remained the same

Build # 26122874472

Build Type

push

github

Committed by

khatchad

Commit Message

Extend precision audit to fixture 59 (`tf.RaggedTensor`)

`tf.RaggedTensor.from_nested_row_splits(values, splits)` — runtime
dtype=int32, shape=(3, None, None). Ariadne emits an INT32 TensorType
with three dims: a known constant 3 followed by two unknown (null)
dims, matching the runtime shape exactly. No precision gap on Layer 1.

Notable Layer-2 behavior: the algorithm replaces null dims with
`SymbolicDim("?")` wildcards in the inferred signature — partial-unknown
per-dim degrades to per-dim wildcards rather than a whole-shape drop.
Documents the graceful-degradation property for ragged-rank inputs.

Inline assertion (not helper-based) because fixtures 60+ that use
`testHasLikelyTensorParameterHelper()` need a separate helper refactor
to carry expected TensorType through to the precision audit.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Coverage Stats

1188 of 1645 relevant lines covered (72.22%)

0.72 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
1	26122874472.1	19 May 2026 08:23PM UTC	31	72.22	GitHub Action Run

ponder-lab / Hybridize-Functions-Refactoring / 26122874472
72%
main: 72%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 26122874472

ponder-lab / Hybridize-Functions-Refactoring / 26122874472 72% main: 72%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 26122874472

ponder-lab / Hybridize-Functions-Refactoring / 26122874472
72%
main: 72%

README BADGES
x