lbotecur commited on
Commit
d4be934
·
verified ·
1 Parent(s): 9fb540d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -6
README.md CHANGED
@@ -10,9 +10,10 @@ tags:
10
 
11
  # TabOmics v2
12
 
13
- Tabular foundation model for wide omics data: classifies the rows of a table in context, in one forward pass,
14
- and gives row embeddings and feature importance. 7.3M parameters; feature clip 10;
15
- column permutation off.
 
16
 
17
  ```python
18
  from tabomics import Classifier
@@ -20,6 +21,4 @@ clf = Classifier(model="v2").fit(X_train, y_train)
20
  proba = clf.predict_proba(X_test)
21
  ```
22
 
23
- Code, documentation and licence: https://github.com/lbotecur/tabomics (GPL-3.0-or-later). The architecture
24
- derives from nanoTabPFN (Pfefferle et al.); the synthetic training tables come from a generator derived from
25
- TabICL's graph prior.
 
10
 
11
  # TabOmics v2
12
 
13
+ TabOmics is a tabular foundation model for wide omics data. It is pretrained on synthetic tables and uses in-context
14
+ learning: given labelled training rows and unlabelled test rows, it predicts the test rows in a single forward pass,
15
+ with no training or tuning on your data. It is designed for wide tables (thousands of features, few samples, as in
16
+ omics), also works on ordinary tables, and provides row embeddings and feature importance.
17
 
18
  ```python
19
  from tabomics import Classifier
 
21
  proba = clf.predict_proba(X_test)
22
  ```
23
 
24
+ Code, documentation and licence: https://github.com/lbotecur/tabomics (GNU General Public License v3.0 or later).