Instructions to use Taykhoom/ERNIE-RNA-SS with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Taykhoom/ERNIE-RNA-SS with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="Taykhoom/ERNIE-RNA-SS", trust_remote_code=True)# Load model directly from transformers import AutoModelForMaskedLM model = AutoModelForMaskedLM.from_pretrained("Taykhoom/ERNIE-RNA-SS", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -47,10 +47,10 @@ and variant effect prediction (`vep-traitgym-mendelian`).
|
|
| 47 |
`bpRNA-new` was excluded: its backbone is identical to the pretrained ERNIE-RNA checkpoint
|
| 48 |
(the backbone was frozen during SS fine-tuning), making it equivalent to `Taykhoom/ERNIE-RNA`.
|
| 49 |
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
|
| 55 |
## Parity Verification
|
| 56 |
|
|
|
|
| 47 |
`bpRNA-new` was excluded: its backbone is identical to the pretrained ERNIE-RNA checkpoint
|
| 48 |
(the backbone was frozen during SS fine-tuning), making it equivalent to `Taykhoom/ERNIE-RNA`.
|
| 49 |
|
| 50 |
+
The remaining five were evaluated via linear probing on three tasks from
|
| 51 |
+
[mRNABench](https://huggingface.co/collections/morrislab/mrnabench): RNA subcellular
|
| 52 |
+
localization, mRNA half-life prediction, and variant effect prediction. `RNA3DB` was the
|
| 53 |
+
best-performing checkpoint across all three tasks and was selected for this release.
|
| 54 |
|
| 55 |
## Parity Verification
|
| 56 |
|