For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Python class

TransformerBlock

TransformerBlock

class max.nn.TransformerBlock(attention, mlp, attention_norm, mlp_norm, residual_multiplier=1.0)

source

Bases: Module

Stack of Attention, FeedForward, and RMSNorm layers.

Parameters:

• attention (Module)
• mlp (Layer)
• attention_norm (Layer)
• mlp_norm (Layer)
• residual_multiplier (float)