metadata
library_name: keras-hub
This is a GPT2 model 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: 24
- num_heads: 16
- hidden_dim: 1024
- intermediate_dim: 4096
- 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 for more information.