fix(export): GPT models w/ bias=False convert properly (#11255)

Signed-off-by: Terry Kong <[email protected]>

nemo/export/trt_llm/tensorrt_llm_build.py

@@ -118,6 +118,14 @@ def build_and_save_engine(
         build_config.lora_config = lora_config
 
     model = model_cls.from_config(model_config)
+    if not model_config.bias and model_config.architecture == 'GPTForCausalLM':
+        # NOTE: GPT models in megatron-core that set bias=False sets the bias false globally
+        #  whereas bias=False in TRTLLM GPT models sets it false everywhere except
+        #  LayerNorm. This change makes TRTLLM's implementation match megatron-core.
+        for name, module in model.named_modules():
+            if isinstance(module, tensorrt_llm.layers.normalization.LayerNorm):
+                module.bias = None
+                module.register_parameter('bias', None)
     model = optimize_model(
         model,
         use_parallel_embedding=model_config.use_parallel_embedding,