TRF011

forward() must not access non-nn.Module attributes on submodules (breaks pipeline parallelism with Identity replacement).

Default	Enabled
Scope	All models
Source	mlinter/trf011.py
Show in terminal	mlinter --rule TRF011

What it does

In forward() methods of PreTrainedModel subclasses, checks for attribute accesses on submodules that would not exist on torch.nn.Identity. This includes attribute accesses on loop variables iterating over self.layers, and self.<submodule>.<attr> chains where <attr> is not a standard nn.Module attribute.

Why is this bad?

Pipeline parallelism may replace any submodule with torch.nn.Identity. Accessing custom attributes (e.g. decoder_layer.attention_type) on a replaced module raises AttributeError at runtime. Per-layer metadata should be read from self.config instead.

Example

 def forward(self, ...):
-    for decoder_layer in self.layers:
+    for i, decoder_layer in enumerate(self.layers):
         hidden_states = decoder_layer(
             hidden_states,
-            attention_mask=causal_mask_mapping[decoder_layer.attention_type],
+            attention_mask=causal_mask_mapping[self.config.layer_types[i]],
         )

Suppressing this rule

Add a # trf-ignore: TRF011 comment on the flagged line or the line directly above it. See Suppressing rules for whole-file directives and when a suppression is the wrong answer.