-
Enabling vectorization (see https://github.com/nod-ai/iree-amd-aie/pull/789) for convolution results in numerical failure. The values are off only slightly (although they are definitely not correct, t…
-
### 🐛 Describe the bug
The vectorization test `vec_test_all_types_AVX2` and `vec_test_all_types_AVX512` has failed cases on division when dtype is complex float or complex double.
Log:
```bash
#…
-
### Search before asking
- [X] I had searched in the [issues](https://github.com/apache/doris/issues?q=is%3Aissue) and found no similar issues.
### Version
Doris: 3.0.2
Hudi: 0.15
### What's Wr…
cgerx updated
3 weeks ago
-
**Reported by mlennert on 26 Feb 2015 16:00 UTC**
It would be nice if r.to.vect allowed to chose only certain category values that should be vectorized. Something in the line of the 'values' parameter…
-
The peak memory footprint is still quite large. More analysis is needed to track down the roots.
-
> * Scalar creation function naming: still no strong opinion, although I would lean towards either `point()`, `linestring()`, etc. or `create_point()`, `create_linestring()`, etc.
>
> Let's open a s…
-
### 🐛 Describe the bug
new_perf_regression in 2024-02-11
| suite | name | batch_size_new | speed_up_new | inductor_new | eager_new | compilation_latency_new | batch_size_old | speed_up_old | induc…
-
Need to include info on how to "monitor things like consumer lag (per topic, or per partition, time would be ideal), consume rate, and produce rate. Going through the metrics currently available on do…
-
Are there any implementations of / plans to implement vectorized point evaluation (and evaluation of the normal, unit_tangent inv_arclength etc.) of Path objects? Accepting numpy arrays in particular …
-
### Description
When we tested the performance of velox, in most cases the HashAggregation performance of velox was much faster than that of presto, but we found that it was slower than presto in o…