Devis Lucato

@clarity99 could you try running these 3 images, see if there's any difference? ``` docker pull --platform linux/arm64 kernelmemory/service:sha-3753853 docker pull --platform linux/arm64 kernelmemory/service:sha-8a2c25d docker pull --platform linux/arm64 kernelmemory/service:sha-70d5f84 ```...

I can see the same error testing in Ubuntu Server 22.04 LTS, using a local VM with UTM on Apple silicon: ``` uname -a Linux ubuntu 5.15.0-92-generic #102-Ubuntu SMP Wed...

As of today, the error is still occurring. Linux ubuntu 5.15.0-92-generic #102-Ubuntu SMP Wed Jan 10 09:37:39 UTC 2024 aarch64 aarch64 aarch64 GNU/Linux sudo docker run --platform linux/amd64 -e OPENAI_API_KEY="xxx"...

I recently reviewed the Dockerfile with support from the Docker team, and looks like there's some error in the use of build platform and target platform. The error can be...

We've refactored the images, to workaround some limitations of the current CI pipeline. The `latest` label is now used only for Intel/AMD CPUs. To run KM docker image on ARM64,...

Update: the `latest` tag now contains both AMD64 and ARM64 images, using docker arch autodetect. As a result the `latest-arm64` has been deprecated.

Posting here some notes from the PR: * KM uses multiple embedding generators, so it's important not to consider **only** the content, but also the generator used and the underlying...

Looks like the PR has become stale, with a few things to address. If this is a pressing problem, the approach should be reusable (e.g. not having to add caching...

yes we need keyed services for LLMs for a bunch of scenarios, for embedding, text generation, chunking etc. Open to look at a PoC but it will take some time...

> this is just what I need to solve current issue I am having, I use LLAMa for chat completion but still want to use OpenAI for embedding; the problem...