Update modeling_minGRULM.py

modeling_minGRULM.py

@@ -46,6 +46,12 @@ class MinGRULMPreTrainedModel(PreTrainedModel):
         elif isinstance(module, nn.LayerNorm):
             module.bias.data.zero_()
             module.weight.data.fill_(1.0)
+        
+        # NaN kontrolü: Tüm parametrelerde NaN varsa, sıfırlama
+        for name, param in module.named_parameters():
+            if torch.isnan(param).any():
+                print(f"NaN detected in parameter {name}. Replacing with zeros.")
+                param.data = torch.nan_to_num(param.data, nan=0.0) 
 
             
 class MinGRULMForCausalLM(PreTrainedModel):