-
**Current Behavior:**
In the current implementation, only points belonging to the worst-performing cluster have their underperforming group scores set to a value less than 1.0. All other points are e…
-
Hi @strengejacke
There are a few problems with ggeffects() of a lme() model
1. Given the latest {insight}, ggeffects sometimes produce errors with lme() if its weights = varFunc() is specified.…
-
**Describe the bug**
The GPU-accelerated implementation from cuml can give **much worse results than the CPU alternative from the package [umap](https://umap-learn.readthedocs.io/en/latest/index.ht…
-
Hello thanks for your project.
i have a question regarding the data for further finetuning, i need to start collecting data, what should be the data characteristics ?
could the data have noise …
-
Hi!
I am trying single-cell data and scDist, and I have some beginner's confusion.
First, does the distance in scDist refer to the difference specified by `fixed.effects` ? If cluster A has a l…
-
### Astronomer Cosmos Version
Other Astronomer Cosmos version (please specify below)
### If "Other Astronomer Cosmos version" selected, which one?
1.4.0
### dbt-core version
1.8.2
### Versions o…
-
After starting the clustering I get this error:
```
[local/evol1][1 shards] map "extract_text" to "('prompt__cluster',)": 100%|█████████████████████████████████████████████████████████████████████…
-
when executing zero shot. the following probs is empty:
topics, probs = topic_model.fit_transform(docs, embeddings)
following configuration:
from umap import UMAP
umap_model = UMAP(n_neighbo…
-
For example, 1M rows x 2K columns, 50 trees of depth 12:
on a single machine - 272 seconds,
on two machines - 447 seconds.
Hardware: 96G RAM, Xeon with 12 cores, 1Gbit/s ethernet network.
Comman…
-
# The issue
> Note, as mentioned in #8790, I think this issue and that one are very similar, with the main difference that I can see being that #8790 is easier to reproduce/simpler to reason through …