amircodota

Large model output is copied when working from a BLS

6

**Description** When querying an onnx model from a python BLS, performance seems to suffer with large outputs. We suspect the output data is being copied. **Triton Information** 22.06 + patch...

I ran into a case where, specifically in inotify, if I'm watching a directory and the user does not have permissions on a subdirectory, the watch will fail. Added a...

proxyReq event fails when behind a corporate proxy

4

We have been using http-proxy for a long time without a problem. (Congrats on that, BTW :-) ) Lately we have been working with a big enterprise customer, were our...

prettytable-rs 0.8.0 fails on rust 1.67 with segfault

Hi, folks Amir from tabnine here. We have recently released support the new for Apple M1 processors. This requires some changes on the plugin side too. We've updated the [how...

linux issue

4

Hi folks, Amir from codota here (the company behind TabNine). We have recently introduced an issue were some linux users might have TabNine stop working. We suspect it might affect...

Memory usage of qdrant on the desktop

2

Hi folks We are using qdrant for a desktop ai assistant. We are running qdrant on the desktop, indexing many documents and allowing to find relevant snippets. Overall we are...

temp_segments grows very quickly

4

We are running into an issue in our production database were all of the sudden disk usage of our database grows very fast until it runs out of space. Looking...

# Environment Knex version: 3.1.0 Database + version: postgres 16.6 OS: linux # Bug I think I'm getting stale connections from the pool with postgres. I have a setup running...