File size: 626 Bytes
5eee00d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
---
library_name: keras-hub
---
This is a [`GPT2` model](https://keras.io/api/keras_hub/models/gpt2) uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
Model config:
* **name:** gpt2_backbone
* **trainable:** True
* **vocabulary_size:** 50257
* **num_layers:** 48
* **num_heads:** 25
* **hidden_dim:** 1600
* **intermediate_dim:** 6400
* **dropout:** 0.1
* **max_sequence_length:** 1024
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.
|