pullrequests Search Results · bucket language:JavaScript language:JavaScript language:JavaScript language:CSS language:Python
Filter by
905k results
!-- pr-sentinel-review:v1 --
PR Sentinel Review
Reviewed commit: 93182c572a731999ea73bcf9a5da7e0923a257ac
The new test file has one wrong assertion. size_bucket(250) returns L , not M , so ...
category:tests_only
size:S
... shape_variants.py had no _placeholder_value helper, so this PR adds it (and
_placeholder_values uses it).
When the backbone s structure depends on the token count, precompile traces it once per bucket ...
ready
!-- claude-projects-attribution: { github_login : manucruzleiva } -- Requested by Manu · project thread
Stacked on #89. Goes with the Stockpile audit PR in cobblemon-ditto-hms. Only the Stockpile section ...
🚀 Summary of Changes
RBLNConfig.decode_batch_bucket_limit is the maximum number of decode batch buckets, but its docstring called it the
largest bucket. This fixes the docstring, which is also the ...
A bucket containing [NaN, -7, 3] currently returns NaN from get_max and get_abs_max with skipna=True, while get_min(...,
skipna=False) returns -7 instead of propagating the NaN.
This keeps NaNs ...
Add bucket_min and bucket_max functions, once the pyresample PR is merged.
- [x] Tests added !-- for all bug fixes or enhancements --
component:resampling
_StreamingDiLoCoFragment._bucketize_and_allreduce can skip pseudogradients when a fragment needs more than one bucket.
If the skipped pseudogradients differ between replicas, the replicas can apply ...
CLA Signed
... to look at. If it isn t the direction you want, I m fine
dropping it or moving to option 2.
throughput_timeseries takes an optional bucket: timedelta, on InsightsService and on both repository ...