update

README.md

@@ -24,11 +24,11 @@ from transformers import AutoModel, CLIPImageProcessor
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
 model = AutoModel.from_pretrained(
-    'jienengchen/ViTamin-L-336px',
+    'jienengchen/ViTamin-L-384px',
     trust_remote_code=True).to(device).eval()
 
 image = Image.open('./image.png').convert('RGB')
-image_processor = CLIPImageProcessor.from_pretrained('jienengchen/ViTamin-L-336px')
+image_processor = CLIPImageProcessor.from_pretrained('jienengchen/ViTamin-L-384px')
 
 pixel_values = image_processor(images=image, return_tensors='pt').pixel_values
 pixel_values = pixel_values.to(torch.bfloat16).cuda()

config.json

@@ -15,7 +15,7 @@
       "timm_proj": "linear",
       "timm_drop": 0.0,
       "timm_drop_path": 0.1,
-      "image_size": 336
+      "image_size": 384
     },
     "text_cfg": {
       "context_length": 77,

preprocessor_config.json

@@ -1,5 +1,5 @@
 {
-    "crop_size": 336,
+    "crop_size": 384,
     "do_center_crop": true,
     "do_normalize": true,
     "do_resize": true,
@@ -15,6 +15,6 @@
         0.27577711
     ],
     "resample": 3,
-    "size": 336
+    "size": 384
   }