-
This Jira added support to interaction pairs to GLM Mojo. However, we are not supporting enum, enum interaction pairs that contains NA’s in them. Support for enum, enum interaction pairs with NAs wi…
-
Relates to https://github.com/h2oai/h2oai/issues/10683 and https://community.mrtrix.org/t/mrtrix-mmap-gpfs-issue/710.
It seems that shared file systems have poor perfomance for mmap.
-
We need a “button” or link to publish a model from a Wave app (running in HAIC) into MLOps. Currently you must download the MOJO from a model trained in a Wave app, and then upload that MOJO file int…
-
As it was already mentioned in [https://h2oai.atlassian.net/browse/PUBDEV-7057|https://h2oai.atlassian.net/browse/PUBDEV-7057|smart-link] there is an argument to be made to include an ONNX converter f…
-
```sh
$ conda install -c h2oai h2o=3.14.0.2
Fetching package metadata ...........
Solving package specifications: .
UnsatisfiableError: The following specifications were found to be in conflic…
-
I have 64 GB of PC RAM.
Which h2oai/h2ogpt-4096-llama2-chat parameter size can I use and I how do I make h2ogpt use the cpu and PC ram instead of gpu for that?
-
There are a number of places that still link to "h2o-dev" and mention 0xdata in H2O 3.0. These should be cleaned up where appropriate.
h2o-dev: https://github.com/h2oai/h2o-3/search?utf8=%E2%9C%…
-
I have several issues with darkmode.
When I've set my windows sytem color mode to "dark", h2oGPT seems to adapt this.
Even though h2oGPT variable `dark: bool = False` is set, it starts up in dark …
-
A nice new 1.8B base model from H2O which would run on a single GPU. Looks like Llama 2 with Mistral tokenizer and might be an easy add: https://huggingface.co/h2oai/h2o-danube2-1.8b-base
rasbt updated
4 months ago
-
## Description
```
2021-07-27 21:58:40,216 C: 8% D:1.8TB M:109.6GB NODE:LOCAL1 10792 WARNING| File "/opt/h2oai/dai/python/lib/python3.8/site-packages/lightgbm_gpu/sklearn.py", line 921,…