Apply Github PR #127 to all HF models

modeling_internvl_chat.py

@@ -43,7 +43,7 @@ class InternVLChatModel(PreTrainedModel):
     def __init__(self, config: InternVLChatConfig, vision_model=None, language_model=None, use_flash_attn=True):
         super().__init__(config)
 
-        assert version_cmp(transformers.__version__, '4.36.2', 'ge')
+        assert version_cmp(transformers.__version__, '4.37.0', 'ge')
         image_size = config.force_image_size or config.vision_config.image_size
         patch_size = config.vision_config.patch_size
         self.patch_size = patch_size
@@ -112,7 +112,7 @@ class InternVLChatModel(PreTrainedModel):
         B, N, C = input_embeds.shape
         input_embeds = input_embeds.reshape(B * N, C)
 
-        if torch.distributed.get_rank() == 0:
+        if torch.distributed.is_initialized() and torch.distributed.get_rank() == 0:
             print(f'dynamic ViT batch size: {vit_batch_size}, images per sample: {vit_batch_size / B}, dynamic token length: {N}')
 
         input_ids = input_ids.reshape(B * N)