deberta_large / README.md
IMvision12's picture
Upload folder using huggingface_hub
4d3595a verified
|
Raw
History Blame Contribute Delete
801 Bytes
metadata
license: mit
library_name: keras
pipeline_tag: fill-mask
tags:
  - keras
  - kerasformers
  - tf
  - jax
  - pytorch
  - backend-agnostic
  - deberta
  - fill-mask
base_model:
  - microsoft/deberta-large

DeBERTa (deberta_large)

A pure Keras 3 port of DeBERTa, converted from microsoft/deberta-large (Hugging Face Transformers).

The weights are backend-agnostic: the same checkpoint loads and runs identically under the JAX, TensorFlow, or PyTorch Keras backend (KERAS_BACKEND=jax|tensorflow|torch), via kerasformers.

Usage

from kerasformers.models.deberta import DebertaModel

model = DebertaModel.from_weights("kerasformers/deberta_large")