Update model

config.json

@@ -9,7 +9,6 @@
   "d_latents": 512,
   "d_model": 704,
   "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
   "id2label": {
     "0": "LABEL_0",
     "1": "LABEL_1",
@@ -1417,6 +1416,7 @@
     "LABEL_99": 99
   },
   "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 2048,
   "model_type": "perceiver",
   "num_blocks": 1,
   "num_cross_attention_heads": 1,
@@ -1424,17 +1424,21 @@
   "num_latents": 784,
   "num_self_attends_per_block": 8,
   "num_self_attention_heads": 8,
-  "position_embedding_init_scale": 0.02,
+  "output_shape": [
+    1,
+    16,
+    224,
+    224
+  ],
   "qk_channels": null,
   "samples_per_patch": 16,
   "self_attention_widening_factor": 1,
-  "seq_len": 2048,
   "torch_dtype": "float32",
   "train_size": [
     368,
     496
   ],
-  "transformers_version": "4.12.0.dev0",
+  "transformers_version": "4.13.0.dev0",
   "use_query_residual": true,
   "v_channels": null,
   "vocab_size": 262

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb45b9ae57204621de8126876ad1c8ddb85ce7b3fb179b8068f01ec7fed90597
-size 79508115
+oid sha256:b4793385536f1a1415efe0f7058bcbdb005d2d653876bb4f914631a423dee70d
+size 79508371