This is a Gemma3 model uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
This model is related to a CausalLM task.
Model config:
name: gemma3_backbone
trainable: True
vocabulary_size: 262144
image_size: None
num_layers: 26
num_query_heads: 4
num_key_value_heads: 1
hidden_dim: 1152
intermediate_dim: 6912
head_dim: 256
query_head_dim_normalize: True
use_query_key_norm: True
use_post_ffw_norm: True
use_post_attention_norm: True
attention_logit_soft_cap: None
final_logit_soft_cap: None
use_sliding_window_attention: True
sliding_window_size: 512
local_rope_scaling_factor: 1.0
global_rope_scaling_factor: 1.0
vision_encoder: None
layer_norm_epsilon: 1e-06
dropout: 0
This model card has been generated automatically and should be completed by the model author. See Model Cards documentation for more information.