Matt Hamann

@naaa760 For dev, you'll want to install `nx` (`npm i -g nx`) and `uv` (`curl -LsSf https://astral.sh/uv/install.sh | sh`). You should then be able to run `nx generate-types`, which should...

@rachmlenig can you update this PR to run e2e tests during PRs? That's an important check before merging new PRs to main. Essentially, we should run all checks on PRs...

> @mhamann mind taking a look particularly at the .py files this changed? I don't want to mess up any of the gguf stuff there are a couple of tweaks...

I refactored a few things to centralize common code. I think this should be good to go. @BobbyRadford any additional comments?

Thanks for creating this issue, @erandeutsch. I'm digging into the underlying cause. Two initial questions: - Can you confirm which version of LlamaFarm you're using? - If you restart the...

@erandeutsch I believe this issue should be resolved in v0.0.20. If you get a chance, upgrade and try again. Let us know!

For now, we can probably use these models + quants as the short list of options. Open to others as well. - unsloth/Qwen3-1.7B-GGUF:Q4_K_M - unsloth/granite-4.0-h-1b-GGUF:Q5_K_M - unsloth/Llama-3.2-1B-Instruct-GGUF:Q5_K_M - unsloth/gpt-oss-20b-GGUF:Q4_K_M -...

Since this is primarily about dealing with _collections_ within a database, it would probably make sense in the future to support multiple collection configs within a single database (or perhaps...

@rgthelen Can we add a small test to the ingest flow to prevent this from regressing in the future?

@rachmlenig can you resolve the merge conflicts on this one? Also, I think it might be good to flip the "noAutoStart" flag in the factory class to `autoStart` being `true`...