-
I'm interested in learning how to use the operators in e2cnn/escnn to implement a function f such that f(Fea_I) = Fea_I_rot. Here, Fea_I = B(I) represents the feature of an image I after passing throu…
-
Currently, CUTLASS only implements a specialization of `atomic_add` for [`half2`](https://github.com/NVIDIA/cutlass/blob/main/include/cutlass/functional.h#L613), but not `nv_bfloat162`. This in turn l…
-
Should support SG, SNMG, and MNMG training on the taobao dataset.
-
**Is your feature request related to a problem? Please describe.**
In cudf-polars, predicate pushdown can result in arbitrary expressions being part of the parquet read phase. Not all of these expres…
-
Filtered search for a more general type of label data is needed for many use cases. Typically, this takes the form of a vector of integers for every vector in the dataset and query. The general predic…
-
**Is your feature request related to a problem? Please describe.**
pylibcudf was initially developed from within cudf Python as a subpackage. As a result, both [its API documentation](https://docs.rap…
vyasr updated
12 hours ago
-
Add DiskANN search functionality for the current Vamana build. Already have a GreedySearch kernel as part of build, but need to edit, optimize, and possibly add more options.
If possible, would be go…
-
### Is this a new feature, an improvement, or a change to existing functionality?
Improvement
### How would you describe the priority of this feature request
Critical (currently preventing usage)
…
-
During porting the ANN benchmarks from RAFT (#130), a list of potential improvements was created:
- [ ] clang-tidy: cleanup naming conventions
- [ ] cleanup/unify logging (including CPU-only builds)…
-
## Design
Kudo serialization format is optimized for columnar batch serialization used during spark shuffle, which significantly improved serialization/deserialization time compared to jcudf serial…