File size: 728 Bytes
90f0009 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
---
library_name: keras-hub
---
This is a [`Albert` model](https://keras.io/api/keras_hub/models/albert) uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
Model config:
* **name:** albert_backbone
* **trainable:** True
* **vocabulary_size:** 30000
* **num_layers:** 12
* **num_heads:** 64
* **num_groups:** 1
* **num_inner_repetitions:** 1
* **embedding_dim:** 128
* **hidden_dim:** 4096
* **intermediate_dim:** 16384
* **dropout:** 0
* **max_sequence_length:** 512
* **num_segments:** 2
This model card has been generated automatically and should be completed by the model author. See [Model Cards documentation](https://huggingface.co/docs/hub/model-cards) for more information.
|