Separate config into text and vision configs

config.json

@@ -1,9 +1,5 @@
 {
     "drop_rate":0.1,
-    "freeze_RoBERTa":false,
-    "freeze_ViT":false,
-    "freeze_layer_count_roberta":false,
-    "freeze_layer_count_vit":false,
     "head_hidden_scale":2,
     "hidden_act":"gelu",
     "hidden_size":768,
@@ -18,23 +14,43 @@
     "mlp_ratio":4,
     "num_attention_heads":12,
     "num_hidden_layers":6,
-    "resolution_before":224,
     "stop_gradient":false,
     "tie_word_embeddings":false,
-    "tokenizer":"roberta-base",
-    "unfreeze_RoBERTa_attention":false,
-    "unfreeze_RoBERTa_embeddings":false,
-    "unfreeze_RoBERTa_encoder":false,
-    "unfreeze_RoBERTa_layernorm":false,
-    "unfreeze_ViT_attention":false,
-    "unfreeze_ViT_layernorm":false,
-    "vit_embed_dim":512,
-    "vit_layers":12,
-    "vit_layernorm_init_from_vit":false,
-    "vit_layernorm_shared":true,
-    "vit_patch_size":16,
-    "vit_remove_last":false,
-    "vit_transformer_width":512,
-    "vit_width":768,
-    "vocab_size":50265
+    "vocab_size":50265,
+    "text_config_dict": null,
+    "text_config":{
+      "architectures": ["BridgeTowerTextModel"],
+      "classifier_dropout": null,
+      "vocab_size": 50265,
+      "hidden_size": 768,
+      "num_hidden_layers": 12,
+      "num_attention_heads": 12,
+      "intermediate_size": 3072,
+      "hidden_act": "gelu",
+      "hidden_dropout_prob": 0.1,
+      "attention_probs_dropout_prob": 0.1,
+      "max_position_embeddings": 514,
+      "type_vocab_size": 1,
+      "initializer_range": 0.02,
+      "layer_norm_eps": 1e-05,
+      "pad_token_id": 1,
+      "bos_token_id": 0,
+      "eos_token_id": 2,
+      "position_embedding_type": "absolute",
+      "use_cache": true
+    },
+    "vision_config_dict": null,
+    "vision_config":{
+      "architectures": ["BridgeTowerVisionModel"],
+      "embed_dim": 512,
+      "input_resolution": 224,
+      "width": 768,
+      "layers": 12,
+      "patch_size": 16,
+      "transformer_width": 512,
+      "output_resolution": 288,
+      "stop_gradient": false,
+      "vit_layernorm_shared": true,
+      "vit_remove_last": false
+    }
 }