Upload SwinForImageClassification

README.md

@@ -12,19 +12,19 @@ model-index:
 - name: cub-200-bird-classifier-swin
   results:
   - task:
-      name: Image Classification
       type: image-classification
+      name: Image Classification
     dataset:
       name: cub-200-2011
       type: cub-200-2011
       args: default
     metrics:
-    - name: validation_accuracy
-      type: accuracy
+    - type: accuracy
       value: 0.8653
-    - name: test_accuracy
-      type: accuracy
+      name: validation_accuracy
+    - type: accuracy
       value: 0.8795
+      name: test_accuracy
 ---
 
 # Model Card for Model ID

config.json

@@ -1,13 +1,21 @@
 {
-  "_name_or_path": "google/vit-base-patch16-224-in21k",
+  "_name_or_path": "microsoft/swin-large-patch4-window12-384-in22k",
   "architectures": [
-    "ViTForImageClassification"
+    "SwinForImageClassification"
   ],
   "attention_probs_dropout_prob": 0.0,
-  "encoder_stride": 16,
+  "depths": [
+    2,
+    2,
+    18,
+    2
+  ],
+  "drop_path_rate": 0.1,
+  "embed_dim": 192,
+  "encoder_stride": 32,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.0,
-  "hidden_size": 768,
+  "hidden_size": 1536,
   "id2label": {
     "0": "LABEL_0",
     "1": "LABEL_1",
@@ -210,9 +218,8 @@
     "198": "LABEL_198",
     "199": "LABEL_199"
   },
-  "image_size": 224,
+  "image_size": 384,
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
   "label2id": {
     "LABEL_0": 0,
     "LABEL_1": 1,
@@ -415,13 +422,35 @@
     "LABEL_98": 98,
     "LABEL_99": 99
   },
-  "layer_norm_eps": 1e-12,
-  "model_type": "vit",
-  "num_attention_heads": 12,
+  "layer_norm_eps": 1e-05,
+  "mlp_ratio": 4.0,
+  "model_type": "swin",
   "num_channels": 3,
-  "num_hidden_layers": 12,
-  "patch_size": 16,
+  "num_heads": [
+    6,
+    12,
+    24,
+    48
+  ],
+  "num_layers": 4,
+  "out_features": [
+    "stage4"
+  ],
+  "out_indices": [
+    4
+  ],
+  "patch_size": 4,
+  "path_norm": true,
   "qkv_bias": true,
+  "stage_names": [
+    "stem",
+    "stage1",
+    "stage2",
+    "stage3",
+    "stage4"
+  ],
   "torch_dtype": "float32",
-  "transformers_version": "4.47.0"
+  "transformers_version": "4.47.0",
+  "use_absolute_embeddings": false,
+  "window_size": 12
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a8aaa29a64b49a90a1dba0dc4c821cbe8fd2a8bf9e4a3c1e352a51004b3b7d41
-size 343833032
+oid sha256:84f32ba9b2450a6d0b91fcf983bf1a2b32b4fb7066f2119bb675887d5d0b9448
+size 786062056