fix vision_config

see https://github.com/OSU-NLP-Group/UGround/issues/12

config.json

@@ -35,9 +35,16 @@
   "use_sliding_window": false,
   "video_token_id": 151656,
   "vision_config": {
+    "depth": 32,
+    "embed_dim": 1280,
+    "mlp_ratio": 4,
+    "num_heads": 16,
     "in_chans": 3,
-    "model_type": "qwen2_vl",
-    "spatial_patch_size": 14
+    "hidden_size": 3584,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2
   },
   "vision_end_token_id": 151653,
   "vision_start_token_id": 151652,