fixed bugs

build_mlp.py

@@ -84,7 +84,7 @@ class CLIPVisionTower(nn.Module):
             pos_tokens = pos_embed_checkpoint[:, num_extra_tokens:]
             pos_tokens = pos_tokens.reshape(-1, orig_size, orig_size,
                                             embedding_size).permute(
-                                                0, 3, 1, 2)
+                                                0, 3, 1, 2).float()
             pos_tokens = torch.nn.functional.interpolate(
                 pos_tokens,
                 size=(new_size, new_size),

modeling_internlm_xcomposer2.py

@@ -483,9 +483,9 @@ class InternLMXComposer2ForCausalLM(InternLM2PreTrainedModel):
         top_p: float = 0.8,
         repetition_penalty: float=1.005,
         meta_instruction:
-        str = 'You are an AI assistant whose name is InternLM-XComposer (浦语·灵笔).\n'
-        '- InternLM-XComposer (浦语·灵笔) is a conversational language model that is developed by Shanghai AI Laboratory (上海人工智能实验室). It is designed to be helpful, honest, and harmless.\n'
-        '- InternLM-XComposer (浦语·灵笔) can understand and communicate fluently in the language chosen by the user such as English and 中文.',
+        str = 'You are an AI assistant whose name is InternLM-XComposer.\n'
+        '- InternLM-XComposer is a conversational language model that is developed by Shanghai AI Laboratory. It is designed to be helpful, honest, and harmless.\n'
+        '- InternLM-XComposer can understand and communicate fluently in English. You must respond only in English.',
         **kwargs,
     ):
         if image is None: