-
This is a proposal on how to approach distribution of an MLIR program across multiple homogeneous physical devices that have peered memory. It is intended to improve memory usage under NUMA. Devices/Q…
-
We need to define how to best define peak annotations, keeping in mind that these need to work for profile and centroided spectra. The current [test implementation](https://github.com/lgatto/Spectrum/…
-
@bryanconnor would you check out this plugin:
http://danielcebrian.com/share-annotation/demo.html
And offer some feedback?
-
hello! thanks for the libraries, i'd be stuffed if i tried doing any of this without them! i am trying to train coco to recognize a new single class, so i made a dataset with polygon annotations and s…
-
## Describe the bug
Fails to parse type of optional `num_iteration` parameter for LightGBM's predict method
```
INFO:[2024-10-24 10:47:23.863] - Analyzing arguments in .
WARNING:[2024-10-24 10:4…
-
When table.apply() is called on a view that has a foreign key, it will attempt to create the existing foreign key. This will fail (because the fkey exists). It also seems to retry forever (this might …
-
Currently all resampling proceeds in ignorance of image acquisition parameters. For designs involving sparse acquisition, post-convolution downsampling to the TR needs to integrate over the acquisitio…
-
When rapidly ingesting with quantization turned on, the full vectors seem to be put into the cache such that the cluster uses significantly more memory than one would expect.
## Current Behavior
…
-
```
\VITS-fast-fine-tuning>python finetune_speaker_v2.py -m ./OUTPUT_MODEL --max_epochs 1000 --drop_speaker_embed True
INFO:OUTPUT_MODEL:{'train': {'log_interval': 10, 'eval_interval': 100, 'see…
-
A+ style guides
https://apluslms.github.io/contribute/styleguides/
https://github.com/google/eng-practices/blob/master/review/reviewer/index.md