Add downstream eval to puzzletron v2 - #2104

Merged

j-rausch merged 16 commits into

feature/puzzletron_v2from

gkarch/add_downstream_eval

Aug 11, 2026

Commits on Aug 11, 2026

added downstream evaluation

grzegorz-k-karch
authored and
j-rausch
committed
fbce1d1
added yaml for nano eval

grzegorz-k-karch
authored and
j-rausch
committed
ff46317
Label downstream evaluation dashboard stages

grzegorz-k-karch
authored and
j-rausch
committed
701dad9
nano/smoke pipeline works with downstream eval

grzegorz-k-karch
authored and
j-rausch
committed
179e521
📝 Add docstrings to `gkarch/add_downstream_eval` (#2105)

coderabbitai[bot]
authored and
j-rausch
committed
d135c1f
fix: CodeRabbit auto-fixes for PR #2104 (#2106)

authored and
j-rausch
committed
5983fd6
drop docstring churn

grzegorz-k-karch
authored and
j-rausch
committed
59bcf30
undid changes not related to evaluation

grzegorz-k-karch
authored and
j-rausch
committed
8ec08ef
fix: validate lmms-eval task completion

grzegorz-k-karch
authored and
j-rausch
committed
63c9f29
fix: align lmms metric suggestions

grzegorz-k-karch
authored and
j-rausch
committed
7eda140
fix: guard lmms-eval reserved args

grzegorz-k-karch
authored and
j-rausch
committed
db5feb5
fix: clean up lmms-eval process group

grzegorz-k-karch
authored and
j-rausch
committed
bdb20f7
docs: pin lmms-eval example version

grzegorz-k-karch
authored and
j-rausch
committed
28aed98
fix: isolate lmms-eval dependency pin

grzegorz-k-karch
authored and
j-rausch
committed
5c8f36b
Fix lmms-eval timeout typing
j-rausch
committed
f15c992
Address downstream evaluation review findings
j-rausch
committed
ebfe4e7