Fix remaining Transformers v5 crash: guard llm_config and to_dict() for None (follow-up to `e980c02`)

configuration_internvl_chat.py

@@ -51,16 +51,19 @@ class InternVLChatConfig(PretrainedConfig):
             self.vision_config = InternVisionPatchConfig(**vision_config)
         else:
             self.vision_config = None
-        if llm_config['architectures'][0] == 'LlamaForCausalLM':
+
+        arch = (llm_config.get("architectures") or [None])[0]
+        if arch == 'LlamaForCausalLM':
             self.llm_config = LlamaConfig(**llm_config)
-        elif llm_config['architectures'][0] == 'InternLM2ForCausalLM':
+        elif arch == 'InternLM2ForCausalLM':
             self.llm_config = InternLM2Config(**llm_config)
-        elif llm_config['architectures'][0] == 'InternLM2VEForCausalLM':
+        elif arch == 'InternLM2VEForCausalLM':
             self.llm_config = InternLM2Config(**llm_config)
-        elif llm_config['architectures'][0] == 'Qwen2ForCausalLM':
+        elif arch == 'Qwen2ForCausalLM':
             self.llm_config = Qwen2Config(**llm_config)
         else:
-            raise ValueError('Unsupported architecture: {}'.format(llm_config['architectures'][0]))
+            self.llm_config = None
+
         self.use_backbone_lora = use_backbone_lora
         self.use_llm_lora = use_llm_lora
         self.pad2square = pad2square
@@ -87,8 +90,8 @@ class InternVLChatConfig(PretrainedConfig):
             `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
         """
         output = copy.deepcopy(self.__dict__)
-        output['vision_config'] = self.vision_config.to_dict()
-        output['llm_config'] = self.llm_config.to_dict()
+        output['vision_config'] = self.vision_config.to_dict() if self.vision_config is not None else {}
+        output['llm_config'] = self.llm_config.to_dict() if self.llm_config is not None else {}
         output['model_type'] = self.__class__.model_type
         output['use_backbone_lora'] = self.use_backbone_lora
         output['use_llm_lora'] = self.use_llm_lora