Mirrors_Framework/deepmind-research
1
0
Fork 0
mirror of https://github.com/google-deepmind/deepmind-research.git synced 2026-05-31 13:05:40 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Clarify the input sequence usage.

Browse Source 
PiperOrigin-RevId: 424333047
This commit is contained in:
Ziga Avsec
2022-01-26 14:39:37 +00:00
committed by Diego de Las Casas
parent cdd3bdb636
commit 7b427b5161
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
enformer/README.md
+2 -1
View File
@@ -80,7 +80,8 @@ The simplest way to perform inference is to load the model via tfhub.dev (TODO:
LINK). The input sequence length is 393,216 with the prediction corresponding to LINK). The input sequence length is 393,216 with the prediction corresponding to
128 base pair windows of the center 114,688 base pairs. The input sequence is 128 base pair windows of the center 114,688 base pairs. The input sequence is
one hot encoded using the order of indices being 'ACGT' with N values being all one hot encoded using the order of indices being 'ACGT' with N values being all
zeros. zeros. Note that only the central 196,608 bp of the input sequence will be used
by the Enformer model. The rest will be cropped within the model.
```python ```python
import tensorflow as tf import tensorflow as tf