Uche Ogbuji
Uche Ogbuji
### Issue you'd like to raise. LC Release announcements seem to be missing from Discord's Announcements channel since 0.0.166. Looking more closely, these seem to be manual, added by hwchase17....
It's been annoying me for a while that each of the metadata fields for our vector DB are a bit different: The ones in `pgvector_data_doc` use: ```sql tags TEXT[] --...
* #78 * #79
Expand the test suite, reaching for 100% coverage. A notable gap in coverage is `async_helper.py`.
I'm hoping to use MLX-LM to try out inference with [this new LM specialized for Nigerian languages](https://huggingface.co/BeardedMonster/SabiYarn-125M). It's based on Karpathy's nanogpt (they say it's inspired by [GPT-J](https://huggingface.co/docs/transformers/en/model_doc/gptj)'s architecture—I suspect...
The new chunkers, post #30, are now generators, but that does us no good with Qdrant helper `update()` requiring a fixed sequence. Also consider the PG helper `insert_many()` as well,...
Following on from muddled notes in #30, Create a token-aware `text_helper.token_splitter()` class which works at the tokenizer level.
We currently have a pre-commit hook which has served us till now. Recently something changed with ruff options which required some tweaking. In that process, however, I noticed that [there...
In supporting concurrent requests, we won't at first assume concurrent inference capability at the model weight stage. This would require us to mutex the LLMs. We'll want control over the...
Server: ```sh toolio_server --model=mlx-community/Phi-3-mini-128k-instruct-4bit ``` Client: ```sh toolio_request --apibase="http://localhost:8000" --prompt='What is the average airspeed of an unladen swallow?' ``` You can run the above any number of times, but as...