First commit

config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "ViTMAEForPreTraining"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "decoder_hidden_size": 512,
+  "decoder_intermediate_size": 2048,
+  "decoder_num_attention_heads": 16,
+  "decoder_num_hidden_layers": 8,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 1024,
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-12,
+  "mask_ratio": 0.75,
+  "model_type": "vit_mae",
+  "norm_pix_loss": false,
+  "num_attention_heads": 16,
+  "num_channels": 3,
+  "num_hidden_layers": 24,
+  "patch_size": 16,
+  "qkv_bias": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.16.0.dev0"
+}

preprocessor_config.json

@@ -0,0 +1,18 @@
+{
+  "_processor_class": null,
+  "do_normalize": true,
+  "do_resize": true,
+  "feature_extractor_type": "ViTMAEFeatureExtractor",
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "resample": 2,
+  "size": 224
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28a019bd1842a7bdea1cfa9bb390047ad3fab82032cea2045873ca0399ca5a84
+size 1318371301