Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
- ncbi/pubmed
|
|
|
|
| 1 |
+
# ModernBioBERT
|
| 2 |
+
|
| 3 |
+
A modern variant of BioBERT based on ModernBERT.
|
| 4 |
+
We continued the masked language modeling pre-training task for 1.000.000 steps on PubMed abstracts.
|
| 5 |
+
|
| 6 |
+
## Pre-Training Details
|
| 7 |
+
```
|
| 8 |
+
Batchsize: 512
|
| 9 |
+
Learningrate: 1e-4
|
| 10 |
+
Warmupsteps: 500
|
| 11 |
+
Learning Rate Scheduler: Cosine Schedule
|
| 12 |
+
Max. Sequence Length: 512
|
| 13 |
+
Precision: bfloat16
|
| 14 |
+
```
|
| 15 |
+
|
| 16 |
---
|
| 17 |
datasets:
|
| 18 |
- ncbi/pubmed
|