Add preprocessor_config.json (from source model)

preprocessor_config.json

@@ -0,0 +1,8 @@
+{
+  "image_processor_type": "Qwen2VLImageProcessor",
+  "max_pixels": 12845056,
+  "min_pixels": 3136,
+  "patch_size": 14,
+  "temporal_patch_size": 2,
+  "merge_size": 2
+}