Pass model to evaluate_coco() to allow importing.

coco.py

@@ -251,7 +251,7 @@ def build_coco_results(dataset, image_ids, rois, class_ids, scores, masks):
     return results
 
 
-def evaluate_coco(dataset, coco, eval_type="bbox", limit=0):
+def evaluate_coco(model, dataset, coco, eval_type="bbox", limit=0):
     """Runs official COCO evaluation.
     dataset: A Dataset object with valiadtion data
     eval_type: "bbox" or "segm" for bounding box or segmentation evaluation
@@ -409,7 +409,7 @@ if __name__ == '__main__':
         dataset_val.prepare()
 
         # TODO: evaluating on 500 images. Set to 0 to evaluate on all images.
-        evaluate_coco(dataset_val, coco, "bbox", limit=500)
+        evaluate_coco(model, dataset_val, coco, "bbox", limit=500)
     else:
         print("'{}' is not recognized. "
               "Use 'train' or 'evaluate'".format(args.command))