upload modeling_llava_phi.py

modeling_llava_phi.py

@@ -193,44 +193,7 @@ class LlavaPhiForCausalLM(PhiForCausalLM, LlavaMetaForCausalLM):
         return model_inputs
 
 
-    # def prepare_inputs_for_generation(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs):
-    #     '''
-    #     This function is called for each token at inference
-    #     '''
-    #     pdb.set_trace()
-    #     images = kwargs.pop("images", None)
 
-        
-    #     _inputs = super().prepare_inputs_for_generation(
-    #         input_ids, past_key_values=past_key_values, inputs_embeds=inputs_embeds, **kwargs
-    #     )
-    #     if images is not None:
-    #         _inputs['images'] = images
-    #     return _inputs
-
-    # def build_chat_input(self, text, images):
-
-    #     return inputs
-
-    # def chat(self, tokenizer, messages: List[dict], stream=False,
-    #          generation_config: Optional[GenerationConfig]=None):
-    #     generation_config = generation_config or self.generation_config
-    #     input_ids = build_chat_input(self, tokenizer, messages, generation_config.max_new_tokens)
-    #     if stream:
-    #         streamer = TextIterStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
-    #         Thread(target=self.generate, kwargs=dict(
-    #             inputs=input_ids, streamer=streamer,
-    #             generation_config=generation_config,
-    #         )).start()
-    #         return streamer
-    #     else:
-    #         outputs = self.generate(input_ids, generation_config=generation_config)
-    #         response = tokenizer.decode(outputs[0][len(input_ids[0]):], skip_special_tokens=True)
-    #         return response
-
-    # def collate_text_input(self, ):
-    #     pass
-    
 
     def chat(
         self,