Set use_cache to false

README.md

@@ -44,7 +44,7 @@ pip install -r requirements.txt
 
 Then you can enter the directory to run the following command.
 ```python
-from transformers import MllamaForConditionalGeneration, AutoProcessor, AutoConfig
+from transformers import MllamaForConditionalGeneration, AutoProcessor
 import torch
 from PIL import Image
 
@@ -64,17 +64,9 @@ model_name = "intfloat/mmE5-mllama-11b-instruct"
 
 # Load Processor and Model
 processor = AutoProcessor.from_pretrained(model_name)
-processor.tokenizer.padding_side = "right"
-
-config = AutoConfig.from_pretrained(model_name)
-if hasattr(config, 'use_cache'):
-    config.use_cache = False
-config.padding_side = "right"
 model = MllamaForConditionalGeneration.from_pretrained(
-    model_name, config=config, 
-    torch_dtype=torch.bfloat16
+    model_name, torch_dtype=torch.bfloat16
 ).to("cuda")
-model.padding_side = "right"
 model.eval()
 
 # Image + Text -> Text

config.json

@@ -96,7 +96,7 @@
     "torchscript": false,
     "typical_p": 1.0,
     "use_bfloat16": false,
-    "use_cache": true,
+    "use_cache": false,
     "vocab_size": 128256
   },
   "torch_dtype": "bfloat16",