Upload standalone vision encoder: vit-only first commit

open_clip_config.json

@@ -1,16 +1,18 @@
 {
-  "embed_dim": 1280,
-  "vision_cfg": {
-    "layers": 32,
-    "width": 1280,
-    "patch_size": 14,
-    "image_size": 224
-  },
-  "text_cfg": {
-    "context_length": 77,
-    "vocab_size": 49408,
-    "width": 512,
-    "heads": 8,
-    "layers": 12
+  "model_cfg": {
+    "embed_dim": 1280,
+    "vision_cfg": {
+      "layers": 32,
+      "width": 1280,
+      "patch_size": 14,
+      "image_size": 224
+    },
+    "text_cfg": {
+      "context_length": 77,
+      "vocab_size": 49408,
+      "width": 512,
+      "heads": 8,
+      "layers": 12
+    }
   }
 }

open_clip_pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5f3dd1c8686e5ccdabbfa9651a9f202e39cf43bb306219cc22aee01bd376a23
-size 7871096
+oid sha256:b76b5481beca5a364716c07f97726a07c138e862c607aed3fb7cddc3c43e071b
+size 2529737094